In [242]:
import warnings
warnings.filterwarnings("ignore")
import pandas as pd
import scipy.stats as stats
import numpy as np
from pandas.api.types import is_numeric_dtype
import statsmodels.api as sm 
import seaborn as sns
from seaborn_qqplot import QQPlot
from matplotlib import pyplot as plt
from scipy.stats import gamma
from sklearn.preprocessing import StandardScaler
import pickle 
from sklearn.metrics import mean_squared_error
#-------Importing tensorflow libraries-------#
from tensorflow.python.keras.models import Sequential
from tensorflow.python.keras.layers import Dense,Dropout
from tensorflow.python.keras.wrappers.scikit_learn import KerasRegressor
#------split and pipeline libraries------#
from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import KFold
from sklearn.pipeline import Pipeline
#--------------------------------------------------#
from statsmodels.stats.outliers_influence import variance_inflation_factor 
sns.set(color_codes=True)
%matplotlib inline
In [243]:
df  = pd.read_csv("Part- 1,2&3 - Signal.csv")
In [244]:
df.head()
Out[244]:
Parameter 1 Parameter 2 Parameter 3 Parameter 4 Parameter 5 Parameter 6 Parameter 7 Parameter 8 Parameter 9 Parameter 10 Parameter 11 Signal_Strength
0 7.4 0.70 0.00 1.9 0.076 11.0 34.0 0.9978 3.51 0.56 9.4 5
1 7.8 0.88 0.00 2.6 0.098 25.0 67.0 0.9968 3.20 0.68 9.8 5
2 7.8 0.76 0.04 2.3 0.092 15.0 54.0 0.9970 3.26 0.65 9.8 5
3 11.2 0.28 0.56 1.9 0.075 17.0 60.0 0.9980 3.16 0.58 9.8 6
4 7.4 0.70 0.00 1.9 0.076 11.0 34.0 0.9978 3.51 0.56 9.4 5
In [245]:
df.isnull().any()
Out[245]:
Parameter 1        False
Parameter 2        False
Parameter 3        False
Parameter 4        False
Parameter 5        False
Parameter 6        False
Parameter 7        False
Parameter 8        False
Parameter 9        False
Parameter 10       False
Parameter 11       False
Signal_Strength    False
dtype: bool
In [246]:
df.info(verbose=True)

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1599 entries, 0 to 1598
Data columns (total 12 columns):
Parameter 1        1599 non-null float64
Parameter 2        1599 non-null float64
Parameter 3        1599 non-null float64
Parameter 4        1599 non-null float64
Parameter 5        1599 non-null float64
Parameter 6        1599 non-null float64
Parameter 7        1599 non-null float64
Parameter 8        1599 non-null float64
Parameter 9        1599 non-null float64
Parameter 10       1599 non-null float64
Parameter 11       1599 non-null float64
Signal_Strength    1599 non-null int64
dtypes: float64(11), int64(1)
memory usage: 150.0 KB

Univariate Analysis

In [247]:
df.describe()
Out[247]:
Parameter 1 Parameter 2 Parameter 3 Parameter 4 Parameter 5 Parameter 6 Parameter 7 Parameter 8 Parameter 9 Parameter 10 Parameter 11 Signal_Strength
count 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000 1599.000000
mean 8.319637 0.527821 0.270976 2.538806 0.087467 15.874922 46.467792 0.996747 3.311113 0.658149 10.422983 5.636023
std 1.741096 0.179060 0.194801 1.409928 0.047065 10.460157 32.895324 0.001887 0.154386 0.169507 1.065668 0.807569
min 4.600000 0.120000 0.000000 0.900000 0.012000 1.000000 6.000000 0.990070 2.740000 0.330000 8.400000 3.000000
25% 7.100000 0.390000 0.090000 1.900000 0.070000 7.000000 22.000000 0.995600 3.210000 0.550000 9.500000 5.000000
50% 7.900000 0.520000 0.260000 2.200000 0.079000 14.000000 38.000000 0.996750 3.310000 0.620000 10.200000 6.000000
75% 9.200000 0.640000 0.420000 2.600000 0.090000 21.000000 62.000000 0.997835 3.400000 0.730000 11.100000 6.000000
max 15.900000 1.580000 1.000000 15.500000 0.611000 72.000000 289.000000 1.003690 4.010000 2.000000 14.900000 8.000000

Distribution plots for each continous parameters

In [248]:
columns = list(df.columns)
for column in columns:
    plt.figure(figsize=(10, 5))
    sns.distplot(df[column], color = "blue").set_title("Distribution of "+column)

Box plot of each parameter with respect to the signal strength

In [249]:
for column in columns:
    plt.figure(figsize=(10, 5))
    sns.boxplot(x='Signal_Strength',y=column, data = df, hue = 'Signal_Strength')

Calculating median , mode ,skewness and kurtosis

In [250]:
for column in df.columns:
    print(100*"*")
    print("Mean of "+str(column)+"="+str(df[column].mean()))
    print("Median of "+str(column)+"="+str(df[column].median()))
    print("Mode of "+str(column)+"="+str(df[column].mode()[0]))
    print("Skewness in "+str(column)+"="+str(df[column].skew()))
    print("Excess Kurtosis in "+str(column)+"="+str(df[column].kurtosis()))
    print(100*"*")

****************************************************************************************************
Mean of Parameter 1=8.319637273295838
Median of Parameter 1=7.9
Mode of Parameter 1=7.2
Skewness in Parameter 1=0.9827514413284587
Excess Kurtosis in Parameter 1=1.1321433977276252
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 2=0.5278205128205131
Median of Parameter 2=0.52
Mode of Parameter 2=0.6
Skewness in Parameter 2=0.6715925723840199
Excess Kurtosis in Parameter 2=1.2255422501791422
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 3=0.2709756097560964
Median of Parameter 3=0.26
Mode of Parameter 3=0.0
Skewness in Parameter 3=0.3183372952546368
Excess Kurtosis in Parameter 3=-0.7889975153633966
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 4=2.5388055034396517
Median of Parameter 4=2.2
Mode of Parameter 4=2.0
Skewness in Parameter 4=4.54065542590319
Excess Kurtosis in Parameter 4=28.617595424475443
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 5=0.08746654158849257
Median of Parameter 5=0.079
Mode of Parameter 5=0.08
Skewness in Parameter 5=5.680346571971722
Excess Kurtosis in Parameter 5=41.71578724757661
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 6=15.874921826141339
Median of Parameter 6=14.0
Mode of Parameter 6=6.0
Skewness in Parameter 6=1.250567293314441
Excess Kurtosis in Parameter 6=2.023562045840575
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 7=46.46779237023139
Median of Parameter 7=38.0
Mode of Parameter 7=28.0
Skewness in Parameter 7=1.515531257594554
Excess Kurtosis in Parameter 7=3.8098244878645744
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 8=0.9967466791744833
Median of Parameter 8=0.99675
Mode of Parameter 8=0.9972
Skewness in Parameter 8=0.07128766294945525
Excess Kurtosis in Parameter 8=0.9340790654648083
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 9=3.311113195747343
Median of Parameter 9=3.31
Mode of Parameter 9=3.3
Skewness in Parameter 9=0.19368349811284427
Excess Kurtosis in Parameter 9=0.806942508246574
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 10=0.6581488430268921
Median of Parameter 10=0.62
Mode of Parameter 10=0.6
Skewness in Parameter 10=2.4286723536602945
Excess Kurtosis in Parameter 10=11.720250727147674
****************************************************************************************************
****************************************************************************************************
Mean of Parameter 11=10.422983114446502
Median of Parameter 11=10.2
Mode of Parameter 11=9.5
Skewness in Parameter 11=0.8608288069184189
Excess Kurtosis in Parameter 11=0.20002931143836733
****************************************************************************************************
****************************************************************************************************
Mean of Signal_Strength=5.6360225140712945
Median of Signal_Strength=6.0
Mode of Signal_Strength=5
Skewness in Signal_Strength=0.21780157547366327
Excess Kurtosis in Signal_Strength=0.2967081197538759
****************************************************************************************************

Observations:

  • Our dataset has outliers as per the boxplot
  • From the above statistics it can be inferred that Parameters 8 and 9 are normally distributed which can be visually confirmed after we go through the distribution plots of the aforementioned parameters.

    • Multivariate Analysis

    In [59]:
    sns.pairplot(df)
    Out[59]:
    <seaborn.axisgrid.PairGrid at 0x10b30248>

    Observation:We can see some correlation between parameter 1 and parameter 8 but we need to check the correlation statistically

    Let us impute outliers detected through z-score method with median

    In [251]:
    df1 = df.copy()
    In [252]:
    for column in df1.drop(columns=['Signal_Strength']).columns:
    df1[column] = df1[column].apply(lambda x: x if
                        ((x - df1[column].mean())/df1[column].std()) <= 3 
                        else
                        df1[column].median()
                        #np.nan
                       )

    Correlation Test

    Backward elimination method

    In [253]:
    def backwardElimination(x, Y, significance_level, columns):
    numVars = len(x[0])
    for i in range(0, numVars):
        regressor_OLS = sm.OLS(Y, x).fit()
        maxVar = max(regressor_OLS.pvalues)
        if maxVar > significance_level:
            for j in range(0, numVars - i):
                if (regressor_OLS.pvalues[j] == maxVar):
                    x = np.delete(x, j, 1)
                    columns = np.delete(columns, j)
                    
    regressor_OLS.summary()
    return x, columns
    In [254]:
    x_ols = df1.drop(columns=['Signal_Strength']).values
y_ols = df1['Signal_Strength']
significance_level = 0.05
selected_columns = df1.drop(columns=['Signal_Strength']).columns
data, selected_columns = backwardElimination(x_ols, y_ols, significance_level, selected_columns)
    In [255]:
    selected_columns
    Out[255]:
    Index(['Parameter 2', 'Parameter 5', 'Parameter 7', 'Parameter 8',
       'Parameter 9', 'Parameter 10', 'Parameter 11'],
      dtype='object')
    In [256]:
    selected_columns = list(selected_columns)
selected_columns.append('Signal_Strength')
After_backward_elimination = df[selected_columns].copy()
    In [257]:
    plt.figure(figsize=(20, 20))
df_corr = After_backward_elimination.drop(columns=['Signal_Strength']).corr(method='pearson')
ax = sns.heatmap(df_corr, annot=True, cmap='YlGnBu')
bottom, top = ax.get_ylim()
ax.set_ylim(bottom + 0.5, top - 0.5)
    Out[257]:
    (7.0, 0.0)

    Observations:

  • From the backward elimination method we can observe that Parameter 1, Parameter 3, Parameter 4 and Parameter 6 have been dropped due to high correlation.
  • The spearman correlation matrix which we created after removing highly correlated features through backward elimination method do not show any highly correlated feature.
  • We will also use VIF scores to further analyse the correlation in our dataset.

    • Variance Inflation factor

    In [258]:
    X = df.drop(columns=['Signal_Strength'])
vif_data = pd.DataFrame()
vif_data['feature'] = X.columns
vif_data['VIF'] = [variance_inflation_factor(X.values,i) for i in range(len(X.columns))]
    In [259]:
    vif_data
    Out[259]:
    feature VIF
    0 Parameter 1 74.452265
    1 Parameter 2 17.060026
    2 Parameter 3 9.183495
    3 Parameter 4 4.662992
    4 Parameter 5 6.554877
    5 Parameter 6 6.442682
    6 Parameter 7 6.519699
    7 Parameter 8 1479.287209
    8 Parameter 9 1070.967685
    9 Parameter 10 21.590621
    10 Parameter 11 124.394866

    Pairplot after removing outliers

    In [260]:
    sns.pairplot(df1)
    Out[260]:
    <seaborn.axisgrid.PairGrid at 0x87796a08>
    In [261]:
    X = df1.drop(columns=['Signal_Strength'])
vif_data = pd.DataFrame()
vif_data['feature'] = X.columns
vif_data['VIF'] = [variance_inflation_factor(X.values,i) for i in range(len(X.columns))]
    In [262]:
    vif_data
    Out[262]:
    feature VIF
    0 Parameter 1 67.525613
    1 Parameter 2 17.700405
    2 Parameter 3 8.259058
    3 Parameter 4 9.553091
    4 Parameter 5 16.211930
    5 Parameter 6 6.542079
    6 Parameter 7 6.370576
    7 Parameter 8 1331.472877
    8 Parameter 9 960.386214
    9 Parameter 10 29.009573
    10 Parameter 11 133.234695

    Observations:

  • After treating outliers we did not see much improvement in the VIF scores of the parameters.They continue to be very high(>5)
  • In order to solve multicolinearity problem we will use PCA.

    • Standard Scaling

    In [263]:
    scaled = df1.copy()
for column in scaled.drop(columns=['Signal_Strength']).columns:
    scaled[column] = scaled[column].apply(lambda x:
                        (x - scaled[column].mean())/scaled[column].std()
                       )

    Finding the optimal number of principal components that could account for maximum variance in data

    In [264]:
    from sklearn.decomposition import PCA
features = list(scaled.drop(columns=['Signal_Strength']).columns)
pca = PCA(n_components=10)
pca.fit(np.array(scaled[features]))
variance = pca.explained_variance_ratio_
var = np.cumsum(np.round(variance,decimals = 3)*100)
var
    Out[264]:
    array([27.4, 45.1, 59.4, 69.7, 77.9, 84.7, 89.9, 94. , 97. , 99.1])
    In [265]:
    plt.ylabel("% Variance explained")
plt.xlabel("Number of features")
plt.title("PCA analysis")
plt.ylim(20,110)
plt.xlim(0,11)
plt.plot(var)
    Out[265]:
    [<matplotlib.lines.Line2D at 0x8928ebc8>]

    Observations: We could pick n_components = 6 as they account for 90% of variance in the dataset

    In [266]:
    pca = PCA(n_components=6)
principalComponents = pca.fit_transform(np.array(scaled[features]))
pcDf = pd.DataFrame(data = principalComponents
             , columns = ['PC1', 'PC2','PC3','PC4','PC5','PC6'])
    In [267]:
    finalDf = pd.concat([pcDf,scaled[['Signal_Strength']]], axis=1)
    In [269]:
    X = finalDf.drop(columns=['Signal_Strength'])
vif_data = pd.DataFrame()
vif_data['feature'] = X.columns
vif_data['VIF'] = [variance_inflation_factor(X.values,i) for i in range(len(X.columns))]
    In [270]:
    vif_data
    Out[270]:
    feature VIF
    0 PC1 1.0
    1 PC2 1.0
    2 PC3 1.0
    3 PC4 1.0
    4 PC5 1.0
    5 PC6 1.0
    In [271]:
    sns.pairplot(finalDf)
    Out[271]:
    <seaborn.axisgrid.PairGrid at 0x8a523c48>

    No correlation among principal components

    Split the data

    In [304]:
    xpca = finalDf.drop(columns=['Signal_Strength']).values
ypca = finalDf['Signal_Strength'].values
X_train, X_test, y_train, y_test = train_test_split(xpca, ypca,test_size=0.33, random_state=42)

    Training the neural network to create regression model

    In [305]:
    model = Sequential()
model.add(Dense(10, input_dim=6, kernel_initializer='normal', activation='relu'))
model.add(Dropout(0.25))
model.add(Dense(10, activation='relu'))
model.add(Dropout(0.25))
model.add(Dense(1, activation='linear'))
model.summary()

    Model: "sequential_18"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_55 (Dense)             (None, 10)                70        
_________________________________________________________________
dropout_30 (Dropout)         (None, 10)                0         
_________________________________________________________________
dense_56 (Dense)             (None, 10)                110       
_________________________________________________________________
dropout_31 (Dropout)         (None, 10)                0         
_________________________________________________________________
dense_57 (Dense)             (None, 1)                 11        
=================================================================
Total params: 191
Trainable params: 191
Non-trainable params: 0
_________________________________________________________________
    In [306]:
    model.compile(loss='mse', optimizer='adam', metrics=['mse','mae'])
    In [307]:
    history = model.fit(X_train, y_train, epochs=200, batch_size=10,  verbose=1, validation_split=0.2)

    Train on 856 samples, validate on 215 samples
Epoch 1/200
856/856 [==============================] - 1s 707us/sample - loss: 26.6416 - mean_squared_error: 26.6416 - mean_absolute_error: 5.0784 - val_loss: 20.7450 - val_mean_squared_error: 20.7450 - val_mean_absolute_error: 4.4613
Epoch 2/200
856/856 [==============================] - 0s 152us/sample - loss: 14.6958 - mean_squared_error: 14.6958 - mean_absolute_error: 3.5782 - val_loss: 7.6434 - val_mean_squared_error: 7.6434 - val_mean_absolute_error: 2.4951
Epoch 3/200
856/856 [==============================] - 0s 153us/sample - loss: 5.9726 - mean_squared_error: 5.9726 - mean_absolute_error: 2.0134 - val_loss: 2.1598 - val_mean_squared_error: 2.1598 - val_mean_absolute_error: 1.2352
Epoch 4/200
856/856 [==============================] - 0s 155us/sample - loss: 3.9955 - mean_squared_error: 3.9955 - mean_absolute_error: 1.6084 - val_loss: 1.4927 - val_mean_squared_error: 1.4927 - val_mean_absolute_error: 1.0268
Epoch 5/200
856/856 [==============================] - 0s 154us/sample - loss: 3.5079 - mean_squared_error: 3.5079 - mean_absolute_error: 1.5178 - val_loss: 1.0485 - val_mean_squared_error: 1.0485 - val_mean_absolute_error: 0.8394
Epoch 6/200
856/856 [==============================] - 0s 157us/sample - loss: 3.3699 - mean_squared_error: 3.3699 - mean_absolute_error: 1.4761 - val_loss: 0.9247 - val_mean_squared_error: 0.9247 - val_mean_absolute_error: 0.7897
Epoch 7/200
856/856 [==============================] - 0s 151us/sample - loss: 3.1335 - mean_squared_error: 3.1335 - mean_absolute_error: 1.4103 - val_loss: 0.8207 - val_mean_squared_error: 0.8207 - val_mean_absolute_error: 0.7274
Epoch 8/200
856/856 [==============================] - 0s 152us/sample - loss: 3.1021 - mean_squared_error: 3.1021 - mean_absolute_error: 1.4162 - val_loss: 0.5782 - val_mean_squared_error: 0.5782 - val_mean_absolute_error: 0.5825
Epoch 9/200
856/856 [==============================] - 0s 152us/sample - loss: 2.7783 - mean_squared_error: 2.7783 - mean_absolute_error: 1.3228 - val_loss: 0.6557 - val_mean_squared_error: 0.6557 - val_mean_absolute_error: 0.6462
Epoch 10/200
856/856 [==============================] - 0s 153us/sample - loss: 2.8290 - mean_squared_error: 2.8290 - mean_absolute_error: 1.3497 - val_loss: 0.6596 - val_mean_squared_error: 0.6596 - val_mean_absolute_error: 0.6508
Epoch 11/200
856/856 [==============================] - 0s 178us/sample - loss: 2.4912 - mean_squared_error: 2.4912 - mean_absolute_error: 1.2442 - val_loss: 0.5894 - val_mean_squared_error: 0.5894 - val_mean_absolute_error: 0.6055
Epoch 12/200
856/856 [==============================] - 0s 159us/sample - loss: 2.3763 - mean_squared_error: 2.3763 - mean_absolute_error: 1.2252 - val_loss: 0.5144 - val_mean_squared_error: 0.5144 - val_mean_absolute_error: 0.5552
Epoch 13/200
856/856 [==============================] - 0s 158us/sample - loss: 2.4728 - mean_squared_error: 2.4728 - mean_absolute_error: 1.2288 - val_loss: 0.5551 - val_mean_squared_error: 0.5551 - val_mean_absolute_error: 0.5704
Epoch 14/200
856/856 [==============================] - 0s 153us/sample - loss: 2.2382 - mean_squared_error: 2.2382 - mean_absolute_error: 1.1931 - val_loss: 0.4569 - val_mean_squared_error: 0.4569 - val_mean_absolute_error: 0.5213
Epoch 15/200
856/856 [==============================] - 0s 155us/sample - loss: 2.2683 - mean_squared_error: 2.2683 - mean_absolute_error: 1.2047 - val_loss: 0.5003 - val_mean_squared_error: 0.5003 - val_mean_absolute_error: 0.5330
Epoch 16/200
856/856 [==============================] - 0s 153us/sample - loss: 2.1854 - mean_squared_error: 2.1854 - mean_absolute_error: 1.1807 - val_loss: 0.4439 - val_mean_squared_error: 0.4439 - val_mean_absolute_error: 0.5068
Epoch 17/200
856/856 [==============================] - 0s 162us/sample - loss: 2.2764 - mean_squared_error: 2.2764 - mean_absolute_error: 1.2038 - val_loss: 0.4102 - val_mean_squared_error: 0.4102 - val_mean_absolute_error: 0.4913
Epoch 18/200
856/856 [==============================] - 0s 150us/sample - loss: 2.2475 - mean_squared_error: 2.2475 - mean_absolute_error: 1.1967 - val_loss: 0.5532 - val_mean_squared_error: 0.5532 - val_mean_absolute_error: 0.5786
Epoch 19/200
856/856 [==============================] - 0s 150us/sample - loss: 2.1785 - mean_squared_error: 2.1785 - mean_absolute_error: 1.1729 - val_loss: 0.4127 - val_mean_squared_error: 0.4127 - val_mean_absolute_error: 0.4896
Epoch 20/200
856/856 [==============================] - 0s 159us/sample - loss: 2.1451 - mean_squared_error: 2.1451 - mean_absolute_error: 1.1709 - val_loss: 0.4805 - val_mean_squared_error: 0.4805 - val_mean_absolute_error: 0.5318
Epoch 21/200
856/856 [==============================] - 0s 167us/sample - loss: 2.1476 - mean_squared_error: 2.1476 - mean_absolute_error: 1.1759 - val_loss: 0.5513 - val_mean_squared_error: 0.5513 - val_mean_absolute_error: 0.5705
Epoch 22/200
856/856 [==============================] - 0s 152us/sample - loss: 1.8763 - mean_squared_error: 1.8763 - mean_absolute_error: 1.0844 - val_loss: 0.5015 - val_mean_squared_error: 0.5015 - val_mean_absolute_error: 0.5538
Epoch 23/200
856/856 [==============================] - 0s 162us/sample - loss: 2.0451 - mean_squared_error: 2.0451 - mean_absolute_error: 1.1283 - val_loss: 0.4595 - val_mean_squared_error: 0.4595 - val_mean_absolute_error: 0.5269
Epoch 24/200
856/856 [==============================] - 0s 159us/sample - loss: 1.8106 - mean_squared_error: 1.8106 - mean_absolute_error: 1.0761 - val_loss: 0.4381 - val_mean_squared_error: 0.4381 - val_mean_absolute_error: 0.5112
Epoch 25/200
856/856 [==============================] - 0s 155us/sample - loss: 1.8514 - mean_squared_error: 1.8514 - mean_absolute_error: 1.0891 - val_loss: 0.4892 - val_mean_squared_error: 0.4892 - val_mean_absolute_error: 0.5356
Epoch 26/200
856/856 [==============================] - 0s 152us/sample - loss: 2.0447 - mean_squared_error: 2.0447 - mean_absolute_error: 1.1280 - val_loss: 0.5103 - val_mean_squared_error: 0.5103 - val_mean_absolute_error: 0.5531
Epoch 27/200
856/856 [==============================] - 0s 155us/sample - loss: 1.8961 - mean_squared_error: 1.8961 - mean_absolute_error: 1.0948 - val_loss: 0.4215 - val_mean_squared_error: 0.4215 - val_mean_absolute_error: 0.4882
Epoch 28/200
856/856 [==============================] - 0s 169us/sample - loss: 1.7215 - mean_squared_error: 1.7215 - mean_absolute_error: 1.0489 - val_loss: 0.4437 - val_mean_squared_error: 0.4437 - val_mean_absolute_error: 0.5110
Epoch 29/200
856/856 [==============================] - 0s 154us/sample - loss: 1.6555 - mean_squared_error: 1.6555 - mean_absolute_error: 1.0281 - val_loss: 0.4133 - val_mean_squared_error: 0.4133 - val_mean_absolute_error: 0.4929
Epoch 30/200
856/856 [==============================] - 0s 155us/sample - loss: 1.8992 - mean_squared_error: 1.8992 - mean_absolute_error: 1.0797 - val_loss: 0.5089 - val_mean_squared_error: 0.5089 - val_mean_absolute_error: 0.5518
Epoch 31/200
856/856 [==============================] - 0s 155us/sample - loss: 1.5771 - mean_squared_error: 1.5771 - mean_absolute_error: 1.0058 - val_loss: 0.4004 - val_mean_squared_error: 0.4004 - val_mean_absolute_error: 0.4741
Epoch 32/200
856/856 [==============================] - 0s 164us/sample - loss: 1.6944 - mean_squared_error: 1.6944 - mean_absolute_error: 1.0220 - val_loss: 0.4212 - val_mean_squared_error: 0.4212 - val_mean_absolute_error: 0.4917
Epoch 33/200
856/856 [==============================] - 0s 148us/sample - loss: 1.5534 - mean_squared_error: 1.5534 - mean_absolute_error: 0.9894 - val_loss: 0.4081 - val_mean_squared_error: 0.4081 - val_mean_absolute_error: 0.4857
Epoch 34/200
856/856 [==============================] - 0s 155us/sample - loss: 1.5091 - mean_squared_error: 1.5091 - mean_absolute_error: 0.9746 - val_loss: 0.4042 - val_mean_squared_error: 0.4042 - val_mean_absolute_error: 0.4812
Epoch 35/200
856/856 [==============================] - 0s 155us/sample - loss: 1.5923 - mean_squared_error: 1.5923 - mean_absolute_error: 0.9987 - val_loss: 0.4152 - val_mean_squared_error: 0.4152 - val_mean_absolute_error: 0.4842
Epoch 36/200
856/856 [==============================] - 0s 162us/sample - loss: 1.6037 - mean_squared_error: 1.6037 - mean_absolute_error: 0.9964 - val_loss: 0.3838 - val_mean_squared_error: 0.3838 - val_mean_absolute_error: 0.4648
Epoch 37/200
856/856 [==============================] - 0s 150us/sample - loss: 1.4778 - mean_squared_error: 1.4778 - mean_absolute_error: 0.9428 - val_loss: 0.3943 - val_mean_squared_error: 0.3943 - val_mean_absolute_error: 0.4747
Epoch 38/200
856/856 [==============================] - 0s 138us/sample - loss: 1.4798 - mean_squared_error: 1.4798 - mean_absolute_error: 0.9551 - val_loss: 0.4481 - val_mean_squared_error: 0.4481 - val_mean_absolute_error: 0.5135
Epoch 39/200
856/856 [==============================] - 0s 148us/sample - loss: 1.4415 - mean_squared_error: 1.4415 - mean_absolute_error: 0.9435 - val_loss: 0.4416 - val_mean_squared_error: 0.4416 - val_mean_absolute_error: 0.4973
Epoch 40/200
856/856 [==============================] - 0s 139us/sample - loss: 1.4665 - mean_squared_error: 1.4665 - mean_absolute_error: 0.9580 - val_loss: 0.4087 - val_mean_squared_error: 0.4087 - val_mean_absolute_error: 0.4825
Epoch 41/200
856/856 [==============================] - 0s 137us/sample - loss: 1.4795 - mean_squared_error: 1.4795 - mean_absolute_error: 0.9752 - val_loss: 0.4107 - val_mean_squared_error: 0.4107 - val_mean_absolute_error: 0.4874
Epoch 42/200
856/856 [==============================] - 0s 146us/sample - loss: 1.4016 - mean_squared_error: 1.4016 - mean_absolute_error: 0.9315 - val_loss: 0.3947 - val_mean_squared_error: 0.3947 - val_mean_absolute_error: 0.4763
Epoch 43/200
856/856 [==============================] - 0s 154us/sample - loss: 1.3179 - mean_squared_error: 1.3179 - mean_absolute_error: 0.9140 - val_loss: 0.4498 - val_mean_squared_error: 0.4498 - val_mean_absolute_error: 0.5099
Epoch 44/200
856/856 [==============================] - 0s 147us/sample - loss: 1.4898 - mean_squared_error: 1.4898 - mean_absolute_error: 0.9573 - val_loss: 0.3819 - val_mean_squared_error: 0.3819 - val_mean_absolute_error: 0.4605
Epoch 45/200
856/856 [==============================] - 0s 147us/sample - loss: 1.2912 - mean_squared_error: 1.2912 - mean_absolute_error: 0.9030 - val_loss: 0.4024 - val_mean_squared_error: 0.4024 - val_mean_absolute_error: 0.4775
Epoch 46/200
856/856 [==============================] - 0s 148us/sample - loss: 1.2445 - mean_squared_error: 1.2445 - mean_absolute_error: 0.8843 - val_loss: 0.4283 - val_mean_squared_error: 0.4283 - val_mean_absolute_error: 0.4954
Epoch 47/200
856/856 [==============================] - 0s 150us/sample - loss: 1.3160 - mean_squared_error: 1.3160 - mean_absolute_error: 0.9027 - val_loss: 0.3856 - val_mean_squared_error: 0.3856 - val_mean_absolute_error: 0.4654
Epoch 48/200
856/856 [==============================] - 0s 148us/sample - loss: 1.3710 - mean_squared_error: 1.3710 - mean_absolute_error: 0.9425 - val_loss: 0.3783 - val_mean_squared_error: 0.3783 - val_mean_absolute_error: 0.4569
Epoch 49/200
856/856 [==============================] - 0s 172us/sample - loss: 1.3440 - mean_squared_error: 1.3440 - mean_absolute_error: 0.9173 - val_loss: 0.4005 - val_mean_squared_error: 0.4005 - val_mean_absolute_error: 0.4715
Epoch 50/200
856/856 [==============================] - 0s 155us/sample - loss: 1.2856 - mean_squared_error: 1.2856 - mean_absolute_error: 0.8909 - val_loss: 0.3680 - val_mean_squared_error: 0.3680 - val_mean_absolute_error: 0.4483
Epoch 51/200
856/856 [==============================] - 0s 137us/sample - loss: 1.2824 - mean_squared_error: 1.2824 - mean_absolute_error: 0.8905 - val_loss: 0.3677 - val_mean_squared_error: 0.3677 - val_mean_absolute_error: 0.4505
Epoch 52/200
856/856 [==============================] - 0s 136us/sample - loss: 1.3016 - mean_squared_error: 1.3016 - mean_absolute_error: 0.8818 - val_loss: 0.3776 - val_mean_squared_error: 0.3776 - val_mean_absolute_error: 0.4561
Epoch 53/200
856/856 [==============================] - 0s 145us/sample - loss: 1.3839 - mean_squared_error: 1.3839 - mean_absolute_error: 0.9292 - val_loss: 0.4095 - val_mean_squared_error: 0.4095 - val_mean_absolute_error: 0.4743
Epoch 54/200
856/856 [==============================] - 0s 151us/sample - loss: 1.2882 - mean_squared_error: 1.2882 - mean_absolute_error: 0.8993 - val_loss: 0.3991 - val_mean_squared_error: 0.3991 - val_mean_absolute_error: 0.4744
Epoch 55/200
856/856 [==============================] - 0s 154us/sample - loss: 1.3144 - mean_squared_error: 1.3144 - mean_absolute_error: 0.9034 - val_loss: 0.3831 - val_mean_squared_error: 0.3831 - val_mean_absolute_error: 0.4613
Epoch 56/200
856/856 [==============================] - 0s 147us/sample - loss: 1.2197 - mean_squared_error: 1.2197 - mean_absolute_error: 0.8806 - val_loss: 0.3678 - val_mean_squared_error: 0.3678 - val_mean_absolute_error: 0.4494
Epoch 57/200
856/856 [==============================] - 0s 152us/sample - loss: 1.2381 - mean_squared_error: 1.2381 - mean_absolute_error: 0.8780 - val_loss: 0.3657 - val_mean_squared_error: 0.3657 - val_mean_absolute_error: 0.4507
Epoch 58/200
856/856 [==============================] - 0s 146us/sample - loss: 1.0967 - mean_squared_error: 1.0967 - mean_absolute_error: 0.8248 - val_loss: 0.4135 - val_mean_squared_error: 0.4135 - val_mean_absolute_error: 0.4777
Epoch 59/200
856/856 [==============================] - 0s 186us/sample - loss: 1.2071 - mean_squared_error: 1.2071 - mean_absolute_error: 0.8616 - val_loss: 0.3840 - val_mean_squared_error: 0.3840 - val_mean_absolute_error: 0.4601
Epoch 60/200
856/856 [==============================] - 0s 169us/sample - loss: 1.2831 - mean_squared_error: 1.2831 - mean_absolute_error: 0.8851 - val_loss: 0.3607 - val_mean_squared_error: 0.3607 - val_mean_absolute_error: 0.4491
Epoch 61/200
856/856 [==============================] - 0s 138us/sample - loss: 1.1248 - mean_squared_error: 1.1248 - mean_absolute_error: 0.8348 - val_loss: 0.3696 - val_mean_squared_error: 0.3696 - val_mean_absolute_error: 0.4585
Epoch 62/200
856/856 [==============================] - 0s 153us/sample - loss: 1.1331 - mean_squared_error: 1.1331 - mean_absolute_error: 0.8356 - val_loss: 0.3796 - val_mean_squared_error: 0.3796 - val_mean_absolute_error: 0.4742
Epoch 63/200
856/856 [==============================] - 0s 140us/sample - loss: 1.0636 - mean_squared_error: 1.0636 - mean_absolute_error: 0.8264 - val_loss: 0.3839 - val_mean_squared_error: 0.3839 - val_mean_absolute_error: 0.4722
Epoch 64/200
856/856 [==============================] - 0s 140us/sample - loss: 1.1082 - mean_squared_error: 1.1082 - mean_absolute_error: 0.8243 - val_loss: 0.3673 - val_mean_squared_error: 0.3673 - val_mean_absolute_error: 0.4637
Epoch 65/200
856/856 [==============================] - 0s 140us/sample - loss: 1.0938 - mean_squared_error: 1.0938 - mean_absolute_error: 0.8188 - val_loss: 0.3737 - val_mean_squared_error: 0.3737 - val_mean_absolute_error: 0.4624
Epoch 66/200
856/856 [==============================] - 0s 136us/sample - loss: 1.0635 - mean_squared_error: 1.0635 - mean_absolute_error: 0.8002 - val_loss: 0.3754 - val_mean_squared_error: 0.3754 - val_mean_absolute_error: 0.4659
Epoch 67/200
856/856 [==============================] - 0s 146us/sample - loss: 1.0004 - mean_squared_error: 1.0004 - mean_absolute_error: 0.7832 - val_loss: 0.3668 - val_mean_squared_error: 0.3668 - val_mean_absolute_error: 0.4605
Epoch 68/200
856/856 [==============================] - 0s 139us/sample - loss: 1.0681 - mean_squared_error: 1.0681 - mean_absolute_error: 0.7971 - val_loss: 0.3724 - val_mean_squared_error: 0.3724 - val_mean_absolute_error: 0.4584
Epoch 69/200
856/856 [==============================] - 0s 145us/sample - loss: 1.0931 - mean_squared_error: 1.0931 - mean_absolute_error: 0.8299 - val_loss: 0.3814 - val_mean_squared_error: 0.3814 - val_mean_absolute_error: 0.4607
Epoch 70/200
856/856 [==============================] - 0s 151us/sample - loss: 0.9624 - mean_squared_error: 0.9624 - mean_absolute_error: 0.7719 - val_loss: 0.3707 - val_mean_squared_error: 0.3707 - val_mean_absolute_error: 0.4597
Epoch 71/200
856/856 [==============================] - 0s 140us/sample - loss: 1.1376 - mean_squared_error: 1.1376 - mean_absolute_error: 0.8279 - val_loss: 0.3850 - val_mean_squared_error: 0.3850 - val_mean_absolute_error: 0.4732
Epoch 72/200
856/856 [==============================] - 0s 139us/sample - loss: 1.0178 - mean_squared_error: 1.0178 - mean_absolute_error: 0.8025 - val_loss: 0.3817 - val_mean_squared_error: 0.3817 - val_mean_absolute_error: 0.4678
Epoch 73/200
856/856 [==============================] - 0s 147us/sample - loss: 0.9649 - mean_squared_error: 0.9649 - mean_absolute_error: 0.7635 - val_loss: 0.4022 - val_mean_squared_error: 0.4022 - val_mean_absolute_error: 0.4810
Epoch 74/200
856/856 [==============================] - 0s 140us/sample - loss: 0.9530 - mean_squared_error: 0.9530 - mean_absolute_error: 0.7577 - val_loss: 0.4002 - val_mean_squared_error: 0.4002 - val_mean_absolute_error: 0.4838
Epoch 75/200
856/856 [==============================] - 0s 145us/sample - loss: 1.0005 - mean_squared_error: 1.0005 - mean_absolute_error: 0.7783 - val_loss: 0.3803 - val_mean_squared_error: 0.3803 - val_mean_absolute_error: 0.4792
Epoch 76/200
856/856 [==============================] - 0s 138us/sample - loss: 1.0332 - mean_squared_error: 1.0332 - mean_absolute_error: 0.7887 - val_loss: 0.3568 - val_mean_squared_error: 0.3568 - val_mean_absolute_error: 0.4648
Epoch 77/200
856/856 [==============================] - 0s 140us/sample - loss: 0.9998 - mean_squared_error: 0.9998 - mean_absolute_error: 0.7829 - val_loss: 0.3663 - val_mean_squared_error: 0.3663 - val_mean_absolute_error: 0.4636
Epoch 78/200
856/856 [==============================] - 0s 152us/sample - loss: 0.9937 - mean_squared_error: 0.9937 - mean_absolute_error: 0.7724 - val_loss: 0.3768 - val_mean_squared_error: 0.3768 - val_mean_absolute_error: 0.4645
Epoch 79/200
856/856 [==============================] - 0s 175us/sample - loss: 0.9502 - mean_squared_error: 0.9502 - mean_absolute_error: 0.7587 - val_loss: 0.3607 - val_mean_squared_error: 0.3607 - val_mean_absolute_error: 0.4539
Epoch 80/200
856/856 [==============================] - 0s 181us/sample - loss: 0.9419 - mean_squared_error: 0.9419 - mean_absolute_error: 0.7558 - val_loss: 0.3705 - val_mean_squared_error: 0.3705 - val_mean_absolute_error: 0.4657
Epoch 81/200
856/856 [==============================] - 0s 150us/sample - loss: 0.9093 - mean_squared_error: 0.9093 - mean_absolute_error: 0.7343 - val_loss: 0.3647 - val_mean_squared_error: 0.3647 - val_mean_absolute_error: 0.4599
Epoch 82/200
856/856 [==============================] - 0s 155us/sample - loss: 0.9077 - mean_squared_error: 0.9077 - mean_absolute_error: 0.7449 - val_loss: 0.3774 - val_mean_squared_error: 0.3774 - val_mean_absolute_error: 0.4597
Epoch 83/200
856/856 [==============================] - 0s 145us/sample - loss: 0.9472 - mean_squared_error: 0.9472 - mean_absolute_error: 0.7589 - val_loss: 0.3863 - val_mean_squared_error: 0.3863 - val_mean_absolute_error: 0.4749
Epoch 84/200
856/856 [==============================] - 0s 190us/sample - loss: 0.9015 - mean_squared_error: 0.9015 - mean_absolute_error: 0.7336 - val_loss: 0.3939 - val_mean_squared_error: 0.3939 - val_mean_absolute_error: 0.4852
Epoch 85/200
856/856 [==============================] - 0s 153us/sample - loss: 0.8564 - mean_squared_error: 0.8564 - mean_absolute_error: 0.7174 - val_loss: 0.3786 - val_mean_squared_error: 0.3786 - val_mean_absolute_error: 0.4748
Epoch 86/200
856/856 [==============================] - 0s 141us/sample - loss: 0.8783 - mean_squared_error: 0.8783 - mean_absolute_error: 0.7252 - val_loss: 0.3879 - val_mean_squared_error: 0.3879 - val_mean_absolute_error: 0.4779
Epoch 87/200
856/856 [==============================] - 0s 145us/sample - loss: 0.8870 - mean_squared_error: 0.8870 - mean_absolute_error: 0.7321 - val_loss: 0.3710 - val_mean_squared_error: 0.3710 - val_mean_absolute_error: 0.4663
Epoch 88/200
856/856 [==============================] - 0s 161us/sample - loss: 0.8600 - mean_squared_error: 0.8600 - mean_absolute_error: 0.7113 - val_loss: 0.3760 - val_mean_squared_error: 0.3760 - val_mean_absolute_error: 0.4750
Epoch 89/200
856/856 [==============================] - 0s 153us/sample - loss: 0.9138 - mean_squared_error: 0.9138 - mean_absolute_error: 0.7583 - val_loss: 0.3877 - val_mean_squared_error: 0.3877 - val_mean_absolute_error: 0.4834
Epoch 90/200
856/856 [==============================] - 0s 145us/sample - loss: 0.8721 - mean_squared_error: 0.8721 - mean_absolute_error: 0.7319 - val_loss: 0.3758 - val_mean_squared_error: 0.3758 - val_mean_absolute_error: 0.4680
Epoch 91/200
856/856 [==============================] - 0s 139us/sample - loss: 0.8789 - mean_squared_error: 0.8789 - mean_absolute_error: 0.7365 - val_loss: 0.4094 - val_mean_squared_error: 0.4094 - val_mean_absolute_error: 0.4984
Epoch 92/200
856/856 [==============================] - 0s 178us/sample - loss: 0.8564 - mean_squared_error: 0.8564 - mean_absolute_error: 0.7342 - val_loss: 0.3923 - val_mean_squared_error: 0.3923 - val_mean_absolute_error: 0.4870
Epoch 93/200
856/856 [==============================] - 0s 171us/sample - loss: 0.8238 - mean_squared_error: 0.8238 - mean_absolute_error: 0.7058 - val_loss: 0.3757 - val_mean_squared_error: 0.3757 - val_mean_absolute_error: 0.4748
Epoch 94/200
856/856 [==============================] - 0s 157us/sample - loss: 0.8626 - mean_squared_error: 0.8626 - mean_absolute_error: 0.7305 - val_loss: 0.3663 - val_mean_squared_error: 0.3663 - val_mean_absolute_error: 0.4693
Epoch 95/200
856/856 [==============================] - 0s 141us/sample - loss: 0.8524 - mean_squared_error: 0.8524 - mean_absolute_error: 0.7340 - val_loss: 0.3798 - val_mean_squared_error: 0.3798 - val_mean_absolute_error: 0.4778
Epoch 96/200
856/856 [==============================] - 0s 136us/sample - loss: 0.8451 - mean_squared_error: 0.8451 - mean_absolute_error: 0.7255 - val_loss: 0.3749 - val_mean_squared_error: 0.3749 - val_mean_absolute_error: 0.4734
Epoch 97/200
856/856 [==============================] - 0s 148us/sample - loss: 0.7934 - mean_squared_error: 0.7934 - mean_absolute_error: 0.7015 - val_loss: 0.3588 - val_mean_squared_error: 0.3588 - val_mean_absolute_error: 0.4664
Epoch 98/200
856/856 [==============================] - 0s 144us/sample - loss: 0.8022 - mean_squared_error: 0.8022 - mean_absolute_error: 0.6961 - val_loss: 0.3587 - val_mean_squared_error: 0.3587 - val_mean_absolute_error: 0.4682
Epoch 99/200
856/856 [==============================] - 0s 136us/sample - loss: 0.7231 - mean_squared_error: 0.7231 - mean_absolute_error: 0.6707 - val_loss: 0.3658 - val_mean_squared_error: 0.3658 - val_mean_absolute_error: 0.4698
Epoch 100/200
856/856 [==============================] - 0s 146us/sample - loss: 0.7715 - mean_squared_error: 0.7715 - mean_absolute_error: 0.6865 - val_loss: 0.3561 - val_mean_squared_error: 0.3561 - val_mean_absolute_error: 0.4710
Epoch 101/200
856/856 [==============================] - 0s 138us/sample - loss: 0.8189 - mean_squared_error: 0.8189 - mean_absolute_error: 0.7081 - val_loss: 0.3673 - val_mean_squared_error: 0.3673 - val_mean_absolute_error: 0.4715
Epoch 102/200
856/856 [==============================] - 0s 148us/sample - loss: 0.7708 - mean_squared_error: 0.7708 - mean_absolute_error: 0.6902 - val_loss: 0.3795 - val_mean_squared_error: 0.3795 - val_mean_absolute_error: 0.4794
Epoch 103/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8192 - mean_squared_error: 0.8192 - mean_absolute_error: 0.7101 - val_loss: 0.3817 - val_mean_squared_error: 0.3817 - val_mean_absolute_error: 0.4792
Epoch 104/200
856/856 [==============================] - 0s 139us/sample - loss: 0.7725 - mean_squared_error: 0.7725 - mean_absolute_error: 0.6933 - val_loss: 0.3649 - val_mean_squared_error: 0.3649 - val_mean_absolute_error: 0.4712
Epoch 105/200
856/856 [==============================] - 0s 141us/sample - loss: 0.7616 - mean_squared_error: 0.7616 - mean_absolute_error: 0.6799 - val_loss: 0.3858 - val_mean_squared_error: 0.3858 - val_mean_absolute_error: 0.4860
Epoch 106/200
856/856 [==============================] - 0s 146us/sample - loss: 0.7531 - mean_squared_error: 0.7531 - mean_absolute_error: 0.6896 - val_loss: 0.3650 - val_mean_squared_error: 0.3650 - val_mean_absolute_error: 0.4796
Epoch 107/200
856/856 [==============================] - 0s 145us/sample - loss: 0.7392 - mean_squared_error: 0.7392 - mean_absolute_error: 0.6716 - val_loss: 0.3858 - val_mean_squared_error: 0.3858 - val_mean_absolute_error: 0.4812
Epoch 108/200
856/856 [==============================] - 0s 147us/sample - loss: 0.7689 - mean_squared_error: 0.7689 - mean_absolute_error: 0.6900 - val_loss: 0.3654 - val_mean_squared_error: 0.3654 - val_mean_absolute_error: 0.4776
Epoch 109/200
856/856 [==============================] - 0s 154us/sample - loss: 0.7857 - mean_squared_error: 0.7857 - mean_absolute_error: 0.6965 - val_loss: 0.3624 - val_mean_squared_error: 0.3624 - val_mean_absolute_error: 0.4734
Epoch 110/200
856/856 [==============================] - 0s 171us/sample - loss: 0.6839 - mean_squared_error: 0.6839 - mean_absolute_error: 0.6500 - val_loss: 0.3553 - val_mean_squared_error: 0.3553 - val_mean_absolute_error: 0.4660
Epoch 111/200
856/856 [==============================] - 0s 182us/sample - loss: 0.7258 - mean_squared_error: 0.7258 - mean_absolute_error: 0.6708 - val_loss: 0.3561 - val_mean_squared_error: 0.3561 - val_mean_absolute_error: 0.4642
Epoch 112/200
856/856 [==============================] - 0s 172us/sample - loss: 0.6828 - mean_squared_error: 0.6828 - mean_absolute_error: 0.6511 - val_loss: 0.3872 - val_mean_squared_error: 0.3872 - val_mean_absolute_error: 0.4867
Epoch 113/200
856/856 [==============================] - 0s 171us/sample - loss: 0.6999 - mean_squared_error: 0.6999 - mean_absolute_error: 0.6528 - val_loss: 0.3711 - val_mean_squared_error: 0.3711 - val_mean_absolute_error: 0.4816
Epoch 114/200
856/856 [==============================] - 0s 157us/sample - loss: 0.7081 - mean_squared_error: 0.7081 - mean_absolute_error: 0.6552 - val_loss: 0.3631 - val_mean_squared_error: 0.3631 - val_mean_absolute_error: 0.4720
Epoch 115/200
856/856 [==============================] - 0s 173us/sample - loss: 0.6375 - mean_squared_error: 0.6375 - mean_absolute_error: 0.6221 - val_loss: 0.3651 - val_mean_squared_error: 0.3651 - val_mean_absolute_error: 0.4752
Epoch 116/200
856/856 [==============================] - 0s 150us/sample - loss: 0.6764 - mean_squared_error: 0.6764 - mean_absolute_error: 0.6516 - val_loss: 0.3731 - val_mean_squared_error: 0.3731 - val_mean_absolute_error: 0.4818
Epoch 117/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6625 - mean_squared_error: 0.6625 - mean_absolute_error: 0.6385 - val_loss: 0.3611 - val_mean_squared_error: 0.3611 - val_mean_absolute_error: 0.4751
Epoch 118/200
856/856 [==============================] - 0s 188us/sample - loss: 0.6790 - mean_squared_error: 0.6790 - mean_absolute_error: 0.6356 - val_loss: 0.3719 - val_mean_squared_error: 0.3719 - val_mean_absolute_error: 0.4794
Epoch 119/200
856/856 [==============================] - 0s 165us/sample - loss: 0.7255 - mean_squared_error: 0.7255 - mean_absolute_error: 0.6637 - val_loss: 0.3860 - val_mean_squared_error: 0.3860 - val_mean_absolute_error: 0.4866
Epoch 120/200
856/856 [==============================] - 0s 145us/sample - loss: 0.7165 - mean_squared_error: 0.7165 - mean_absolute_error: 0.6569 - val_loss: 0.3689 - val_mean_squared_error: 0.3689 - val_mean_absolute_error: 0.4783
Epoch 121/200
856/856 [==============================] - 0s 154us/sample - loss: 0.6922 - mean_squared_error: 0.6922 - mean_absolute_error: 0.6533 - val_loss: 0.3688 - val_mean_squared_error: 0.3688 - val_mean_absolute_error: 0.4722
Epoch 122/200
856/856 [==============================] - 0s 155us/sample - loss: 0.6554 - mean_squared_error: 0.6554 - mean_absolute_error: 0.6355 - val_loss: 0.3702 - val_mean_squared_error: 0.3702 - val_mean_absolute_error: 0.4706
Epoch 123/200
856/856 [==============================] - 0s 140us/sample - loss: 0.7037 - mean_squared_error: 0.7037 - mean_absolute_error: 0.6579 - val_loss: 0.3751 - val_mean_squared_error: 0.3751 - val_mean_absolute_error: 0.4684
Epoch 124/200
856/856 [==============================] - 0s 151us/sample - loss: 0.6516 - mean_squared_error: 0.6516 - mean_absolute_error: 0.6360 - val_loss: 0.3567 - val_mean_squared_error: 0.3567 - val_mean_absolute_error: 0.4707
Epoch 125/200
856/856 [==============================] - 0s 146us/sample - loss: 0.6550 - mean_squared_error: 0.6550 - mean_absolute_error: 0.6336 - val_loss: 0.3637 - val_mean_squared_error: 0.3637 - val_mean_absolute_error: 0.4734
Epoch 126/200
856/856 [==============================] - 0s 166us/sample - loss: 0.6456 - mean_squared_error: 0.6456 - mean_absolute_error: 0.6357 - val_loss: 0.3647 - val_mean_squared_error: 0.3647 - val_mean_absolute_error: 0.4741
Epoch 127/200
856/856 [==============================] - 0s 169us/sample - loss: 0.6719 - mean_squared_error: 0.6719 - mean_absolute_error: 0.6561 - val_loss: 0.3646 - val_mean_squared_error: 0.3646 - val_mean_absolute_error: 0.4775
Epoch 128/200
856/856 [==============================] - 0s 138us/sample - loss: 0.6292 - mean_squared_error: 0.6292 - mean_absolute_error: 0.6148 - val_loss: 0.3740 - val_mean_squared_error: 0.3740 - val_mean_absolute_error: 0.4847
Epoch 129/200
856/856 [==============================] - 0s 159us/sample - loss: 0.6230 - mean_squared_error: 0.6230 - mean_absolute_error: 0.6173 - val_loss: 0.3649 - val_mean_squared_error: 0.3649 - val_mean_absolute_error: 0.4806
Epoch 130/200
856/856 [==============================] - 0s 169us/sample - loss: 0.6460 - mean_squared_error: 0.6460 - mean_absolute_error: 0.6185 - val_loss: 0.3656 - val_mean_squared_error: 0.3656 - val_mean_absolute_error: 0.4786
Epoch 131/200
856/856 [==============================] - 0s 154us/sample - loss: 0.6312 - mean_squared_error: 0.6312 - mean_absolute_error: 0.6196 - val_loss: 0.3707 - val_mean_squared_error: 0.3707 - val_mean_absolute_error: 0.4817
Epoch 132/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6219 - mean_squared_error: 0.6219 - mean_absolute_error: 0.6118 - val_loss: 0.3867 - val_mean_squared_error: 0.3867 - val_mean_absolute_error: 0.4874
Epoch 133/200
856/856 [==============================] - 0s 161us/sample - loss: 0.6144 - mean_squared_error: 0.6144 - mean_absolute_error: 0.6089 - val_loss: 0.3683 - val_mean_squared_error: 0.3683 - val_mean_absolute_error: 0.4793
Epoch 134/200
856/856 [==============================] - 0s 160us/sample - loss: 0.6206 - mean_squared_error: 0.6206 - mean_absolute_error: 0.6209 - val_loss: 0.3623 - val_mean_squared_error: 0.3623 - val_mean_absolute_error: 0.4720
Epoch 135/200
856/856 [==============================] - 0s 151us/sample - loss: 0.5997 - mean_squared_error: 0.5997 - mean_absolute_error: 0.6019 - val_loss: 0.3544 - val_mean_squared_error: 0.3544 - val_mean_absolute_error: 0.4702
Epoch 136/200
856/856 [==============================] - 0s 164us/sample - loss: 0.5888 - mean_squared_error: 0.5888 - mean_absolute_error: 0.6041 - val_loss: 0.3595 - val_mean_squared_error: 0.3595 - val_mean_absolute_error: 0.4729
Epoch 137/200
856/856 [==============================] - 0s 165us/sample - loss: 0.5919 - mean_squared_error: 0.5919 - mean_absolute_error: 0.5987 - val_loss: 0.3608 - val_mean_squared_error: 0.3608 - val_mean_absolute_error: 0.4734
Epoch 138/200
856/856 [==============================] - 0s 151us/sample - loss: 0.5636 - mean_squared_error: 0.5636 - mean_absolute_error: 0.5797 - val_loss: 0.3788 - val_mean_squared_error: 0.3788 - val_mean_absolute_error: 0.4864
Epoch 139/200
856/856 [==============================] - 0s 158us/sample - loss: 0.5977 - mean_squared_error: 0.5977 - mean_absolute_error: 0.6006 - val_loss: 0.3748 - val_mean_squared_error: 0.3748 - val_mean_absolute_error: 0.4814
Epoch 140/200
856/856 [==============================] - 0s 161us/sample - loss: 0.6004 - mean_squared_error: 0.6004 - mean_absolute_error: 0.6120 - val_loss: 0.3638 - val_mean_squared_error: 0.3638 - val_mean_absolute_error: 0.4761
Epoch 141/200
856/856 [==============================] - 0s 159us/sample - loss: 0.6116 - mean_squared_error: 0.6116 - mean_absolute_error: 0.6077 - val_loss: 0.3647 - val_mean_squared_error: 0.3647 - val_mean_absolute_error: 0.4753
Epoch 142/200
856/856 [==============================] - 0s 154us/sample - loss: 0.5804 - mean_squared_error: 0.5804 - mean_absolute_error: 0.5895 - val_loss: 0.3599 - val_mean_squared_error: 0.3599 - val_mean_absolute_error: 0.4749
Epoch 143/200
856/856 [==============================] - 0s 154us/sample - loss: 0.5907 - mean_squared_error: 0.5907 - mean_absolute_error: 0.6013 - val_loss: 0.3730 - val_mean_squared_error: 0.3730 - val_mean_absolute_error: 0.4829
Epoch 144/200
856/856 [==============================] - 0s 158us/sample - loss: 0.5598 - mean_squared_error: 0.5598 - mean_absolute_error: 0.5767 - val_loss: 0.3635 - val_mean_squared_error: 0.3635 - val_mean_absolute_error: 0.4794
Epoch 145/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5963 - mean_squared_error: 0.5963 - mean_absolute_error: 0.5988 - val_loss: 0.3628 - val_mean_squared_error: 0.3628 - val_mean_absolute_error: 0.4758
Epoch 146/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5703 - mean_squared_error: 0.5703 - mean_absolute_error: 0.5910 - val_loss: 0.3549 - val_mean_squared_error: 0.3549 - val_mean_absolute_error: 0.4717
Epoch 147/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5442 - mean_squared_error: 0.5442 - mean_absolute_error: 0.5774 - val_loss: 0.3645 - val_mean_squared_error: 0.3645 - val_mean_absolute_error: 0.4740
Epoch 148/200
856/856 [==============================] - 0s 145us/sample - loss: 0.5716 - mean_squared_error: 0.5716 - mean_absolute_error: 0.5908 - val_loss: 0.3645 - val_mean_squared_error: 0.3645 - val_mean_absolute_error: 0.4781
Epoch 149/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5817 - mean_squared_error: 0.5817 - mean_absolute_error: 0.6010 - val_loss: 0.3619 - val_mean_squared_error: 0.3619 - val_mean_absolute_error: 0.4784
Epoch 150/200
856/856 [==============================] - 0s 138us/sample - loss: 0.5815 - mean_squared_error: 0.5815 - mean_absolute_error: 0.6028 - val_loss: 0.3536 - val_mean_squared_error: 0.3536 - val_mean_absolute_error: 0.4722
Epoch 151/200
856/856 [==============================] - 0s 143us/sample - loss: 0.5714 - mean_squared_error: 0.5714 - mean_absolute_error: 0.5883 - val_loss: 0.3578 - val_mean_squared_error: 0.3578 - val_mean_absolute_error: 0.4738
Epoch 152/200
856/856 [==============================] - 0s 145us/sample - loss: 0.5688 - mean_squared_error: 0.5688 - mean_absolute_error: 0.5847 - val_loss: 0.3553 - val_mean_squared_error: 0.3553 - val_mean_absolute_error: 0.4704
Epoch 153/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5589 - mean_squared_error: 0.5589 - mean_absolute_error: 0.5854 - val_loss: 0.3736 - val_mean_squared_error: 0.3736 - val_mean_absolute_error: 0.4812
Epoch 154/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5790 - mean_squared_error: 0.5790 - mean_absolute_error: 0.5883 - val_loss: 0.3843 - val_mean_squared_error: 0.3843 - val_mean_absolute_error: 0.4889
Epoch 155/200
856/856 [==============================] - 0s 141us/sample - loss: 0.5578 - mean_squared_error: 0.5578 - mean_absolute_error: 0.5794 - val_loss: 0.3538 - val_mean_squared_error: 0.3538 - val_mean_absolute_error: 0.4720
Epoch 156/200
856/856 [==============================] - 0s 151us/sample - loss: 0.5610 - mean_squared_error: 0.5610 - mean_absolute_error: 0.5843 - val_loss: 0.3605 - val_mean_squared_error: 0.3605 - val_mean_absolute_error: 0.4722
Epoch 157/200
856/856 [==============================] - 0s 151us/sample - loss: 0.5144 - mean_squared_error: 0.5144 - mean_absolute_error: 0.5602 - val_loss: 0.3713 - val_mean_squared_error: 0.3713 - val_mean_absolute_error: 0.4811
Epoch 158/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5269 - mean_squared_error: 0.5269 - mean_absolute_error: 0.5670 - val_loss: 0.3542 - val_mean_squared_error: 0.3542 - val_mean_absolute_error: 0.4709
Epoch 159/200
856/856 [==============================] - 0s 144us/sample - loss: 0.5238 - mean_squared_error: 0.5238 - mean_absolute_error: 0.5627 - val_loss: 0.3564 - val_mean_squared_error: 0.3564 - val_mean_absolute_error: 0.4673
Epoch 160/200
856/856 [==============================] - 0s 154us/sample - loss: 0.5158 - mean_squared_error: 0.5158 - mean_absolute_error: 0.5597 - val_loss: 0.3656 - val_mean_squared_error: 0.3656 - val_mean_absolute_error: 0.4765
Epoch 161/200
856/856 [==============================] - 0s 151us/sample - loss: 0.5406 - mean_squared_error: 0.5406 - mean_absolute_error: 0.5681 - val_loss: 0.3546 - val_mean_squared_error: 0.3546 - val_mean_absolute_error: 0.4721
Epoch 162/200
856/856 [==============================] - 0s 172us/sample - loss: 0.5272 - mean_squared_error: 0.5272 - mean_absolute_error: 0.5708 - val_loss: 0.3606 - val_mean_squared_error: 0.3606 - val_mean_absolute_error: 0.4777
Epoch 163/200
856/856 [==============================] - 0s 154us/sample - loss: 0.5283 - mean_squared_error: 0.5283 - mean_absolute_error: 0.5707 - val_loss: 0.3603 - val_mean_squared_error: 0.3603 - val_mean_absolute_error: 0.4750
Epoch 164/200
856/856 [==============================] - 0s 153us/sample - loss: 0.5102 - mean_squared_error: 0.5102 - mean_absolute_error: 0.5588 - val_loss: 0.3702 - val_mean_squared_error: 0.3702 - val_mean_absolute_error: 0.4791
Epoch 165/200
856/856 [==============================] - 0s 185us/sample - loss: 0.5257 - mean_squared_error: 0.5257 - mean_absolute_error: 0.5630 - val_loss: 0.3642 - val_mean_squared_error: 0.3642 - val_mean_absolute_error: 0.4794
Epoch 166/200
856/856 [==============================] - 0s 139us/sample - loss: 0.4972 - mean_squared_error: 0.4972 - mean_absolute_error: 0.5546 - val_loss: 0.3559 - val_mean_squared_error: 0.3559 - val_mean_absolute_error: 0.4712
Epoch 167/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5338 - mean_squared_error: 0.5338 - mean_absolute_error: 0.5655 - val_loss: 0.3598 - val_mean_squared_error: 0.3598 - val_mean_absolute_error: 0.4733
Epoch 168/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5335 - mean_squared_error: 0.5335 - mean_absolute_error: 0.5657 - val_loss: 0.3596 - val_mean_squared_error: 0.3596 - val_mean_absolute_error: 0.4742
Epoch 169/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5251 - mean_squared_error: 0.5251 - mean_absolute_error: 0.5655 - val_loss: 0.3603 - val_mean_squared_error: 0.3603 - val_mean_absolute_error: 0.4723
Epoch 170/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5126 - mean_squared_error: 0.5126 - mean_absolute_error: 0.5653 - val_loss: 0.3614 - val_mean_squared_error: 0.3614 - val_mean_absolute_error: 0.4729
Epoch 171/200
856/856 [==============================] - 0s 155us/sample - loss: 0.5175 - mean_squared_error: 0.5175 - mean_absolute_error: 0.5553 - val_loss: 0.3669 - val_mean_squared_error: 0.3669 - val_mean_absolute_error: 0.4795
Epoch 172/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5388 - mean_squared_error: 0.5388 - mean_absolute_error: 0.5732 - val_loss: 0.3700 - val_mean_squared_error: 0.3700 - val_mean_absolute_error: 0.4832
Epoch 173/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5274 - mean_squared_error: 0.5274 - mean_absolute_error: 0.5668 - val_loss: 0.3590 - val_mean_squared_error: 0.3590 - val_mean_absolute_error: 0.4745
Epoch 174/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5219 - mean_squared_error: 0.5219 - mean_absolute_error: 0.5676 - val_loss: 0.3578 - val_mean_squared_error: 0.3578 - val_mean_absolute_error: 0.4729
Epoch 175/200
856/856 [==============================] - 0s 138us/sample - loss: 0.5514 - mean_squared_error: 0.5514 - mean_absolute_error: 0.5749 - val_loss: 0.3668 - val_mean_squared_error: 0.3668 - val_mean_absolute_error: 0.4812
Epoch 176/200
856/856 [==============================] - 0s 144us/sample - loss: 0.5240 - mean_squared_error: 0.5240 - mean_absolute_error: 0.5750 - val_loss: 0.3601 - val_mean_squared_error: 0.3601 - val_mean_absolute_error: 0.4755
Epoch 177/200
856/856 [==============================] - 0s 145us/sample - loss: 0.5102 - mean_squared_error: 0.5102 - mean_absolute_error: 0.5582 - val_loss: 0.3645 - val_mean_squared_error: 0.3645 - val_mean_absolute_error: 0.4789
Epoch 178/200
856/856 [==============================] - 0s 134us/sample - loss: 0.5091 - mean_squared_error: 0.5091 - mean_absolute_error: 0.5605 - val_loss: 0.3592 - val_mean_squared_error: 0.3592 - val_mean_absolute_error: 0.4763
Epoch 179/200
856/856 [==============================] - 0s 145us/sample - loss: 0.5197 - mean_squared_error: 0.5197 - mean_absolute_error: 0.5715 - val_loss: 0.3603 - val_mean_squared_error: 0.3603 - val_mean_absolute_error: 0.4775
Epoch 180/200
856/856 [==============================] - 0s 145us/sample - loss: 0.4918 - mean_squared_error: 0.4918 - mean_absolute_error: 0.5496 - val_loss: 0.3586 - val_mean_squared_error: 0.3586 - val_mean_absolute_error: 0.4743
Epoch 181/200
856/856 [==============================] - 0s 138us/sample - loss: 0.5166 - mean_squared_error: 0.5166 - mean_absolute_error: 0.5670 - val_loss: 0.3573 - val_mean_squared_error: 0.3573 - val_mean_absolute_error: 0.4728
Epoch 182/200
856/856 [==============================] - 0s 147us/sample - loss: 0.4981 - mean_squared_error: 0.4981 - mean_absolute_error: 0.5531 - val_loss: 0.3543 - val_mean_squared_error: 0.3543 - val_mean_absolute_error: 0.4714
Epoch 183/200
856/856 [==============================] - 0s 152us/sample - loss: 0.4916 - mean_squared_error: 0.4916 - mean_absolute_error: 0.5530 - val_loss: 0.3557 - val_mean_squared_error: 0.3557 - val_mean_absolute_error: 0.4736
Epoch 184/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5209 - mean_squared_error: 0.5209 - mean_absolute_error: 0.5701 - val_loss: 0.3646 - val_mean_squared_error: 0.3646 - val_mean_absolute_error: 0.4801
Epoch 185/200
856/856 [==============================] - 0s 143us/sample - loss: 0.5306 - mean_squared_error: 0.5306 - mean_absolute_error: 0.5620 - val_loss: 0.3590 - val_mean_squared_error: 0.3590 - val_mean_absolute_error: 0.4755
Epoch 186/200
856/856 [==============================] - 0s 154us/sample - loss: 0.4884 - mean_squared_error: 0.4884 - mean_absolute_error: 0.5489 - val_loss: 0.3669 - val_mean_squared_error: 0.3669 - val_mean_absolute_error: 0.4824
Epoch 187/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5013 - mean_squared_error: 0.5013 - mean_absolute_error: 0.5569 - val_loss: 0.3546 - val_mean_squared_error: 0.3546 - val_mean_absolute_error: 0.4722
Epoch 188/200
856/856 [==============================] - 0s 138us/sample - loss: 0.4951 - mean_squared_error: 0.4951 - mean_absolute_error: 0.5482 - val_loss: 0.3637 - val_mean_squared_error: 0.3637 - val_mean_absolute_error: 0.4813
Epoch 189/200
856/856 [==============================] - 0s 162us/sample - loss: 0.5112 - mean_squared_error: 0.5112 - mean_absolute_error: 0.5627 - val_loss: 0.3581 - val_mean_squared_error: 0.3581 - val_mean_absolute_error: 0.4761
Epoch 190/200
856/856 [==============================] - 0s 176us/sample - loss: 0.5144 - mean_squared_error: 0.5144 - mean_absolute_error: 0.5701 - val_loss: 0.3616 - val_mean_squared_error: 0.3616 - val_mean_absolute_error: 0.4772
Epoch 191/200
856/856 [==============================] - 0s 152us/sample - loss: 0.5125 - mean_squared_error: 0.5125 - mean_absolute_error: 0.5607 - val_loss: 0.3730 - val_mean_squared_error: 0.3730 - val_mean_absolute_error: 0.4886
Epoch 192/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5111 - mean_squared_error: 0.5111 - mean_absolute_error: 0.5581 - val_loss: 0.3665 - val_mean_squared_error: 0.3665 - val_mean_absolute_error: 0.4834
Epoch 193/200
856/856 [==============================] - 0s 151us/sample - loss: 0.5090 - mean_squared_error: 0.5090 - mean_absolute_error: 0.5600 - val_loss: 0.3586 - val_mean_squared_error: 0.3586 - val_mean_absolute_error: 0.4763
Epoch 194/200
856/856 [==============================] - 0s 155us/sample - loss: 0.4744 - mean_squared_error: 0.4744 - mean_absolute_error: 0.5444 - val_loss: 0.3598 - val_mean_squared_error: 0.3598 - val_mean_absolute_error: 0.4767
Epoch 195/200
856/856 [==============================] - 0s 155us/sample - loss: 0.5139 - mean_squared_error: 0.5139 - mean_absolute_error: 0.5652 - val_loss: 0.3643 - val_mean_squared_error: 0.3643 - val_mean_absolute_error: 0.4809
Epoch 196/200
856/856 [==============================] - 0s 154us/sample - loss: 0.4856 - mean_squared_error: 0.4856 - mean_absolute_error: 0.5465 - val_loss: 0.3589 - val_mean_squared_error: 0.3589 - val_mean_absolute_error: 0.4763
Epoch 197/200
856/856 [==============================] - 0s 157us/sample - loss: 0.5062 - mean_squared_error: 0.5062 - mean_absolute_error: 0.5616 - val_loss: 0.3602 - val_mean_squared_error: 0.3602 - val_mean_absolute_error: 0.4783
Epoch 198/200
856/856 [==============================] - 0s 162us/sample - loss: 0.4805 - mean_squared_error: 0.4805 - mean_absolute_error: 0.5557 - val_loss: 0.3602 - val_mean_squared_error: 0.3602 - val_mean_absolute_error: 0.4763
Epoch 199/200
856/856 [==============================] - 0s 144us/sample - loss: 0.4877 - mean_squared_error: 0.4877 - mean_absolute_error: 0.5545 - val_loss: 0.3620 - val_mean_squared_error: 0.3620 - val_mean_absolute_error: 0.4774
Epoch 200/200
856/856 [==============================] - 0s 140us/sample - loss: 0.5055 - mean_squared_error: 0.5055 - mean_absolute_error: 0.5624 - val_loss: 0.3683 - val_mean_squared_error: 0.3683 - val_mean_absolute_error: 0.4835
    In [308]:
    print(history.history.keys())
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('model loss')
plt.ylabel('loss')
plt.xlabel('epoch')
plt.legend(['train', 'validation'], loc='upper left')
plt.show()

    dict_keys(['loss', 'mean_squared_error', 'mean_absolute_error', 'val_loss', 'val_mean_squared_error', 'val_mean_absolute_error'])
    In [309]:
    y_pred = model.predict(X_test)
mean_squared_error(y_test, y_pred)
    Out[309]:
    0.46068788382694303

    Observations:

  • We got the mse score of 0.46 when we trained our neural network on the principal components.
  • Next we will try to train our neural network on the dataset that we received after removing correlated features that were detected through backward elimination algorithm

    • Training neural net on dataset obtained after backward elimination

    In [310]:
    for column in After_backward_elimination.drop(columns=['Signal_Strength']).columns:
    After_backward_elimination[column] = After_backward_elimination[column].apply(lambda x:
                        (x - After_backward_elimination[column].mean())/After_backward_elimination[column].std()
                       )
    In [311]:
    After_backward_elimination
    Out[311]:
    Parameter 2 Parameter 5 Parameter 7 Parameter 8 Parameter 9 Parameter 10 Parameter 11 Signal_Strength
    0 0.961576 -0.243630 -0.379014 0.558100 1.288240 -0.579025 -0.959946 5
    1 1.966827 0.223805 0.624168 0.028252 -0.719708 0.128910 -0.584594 5
    2 1.296660 0.096323 0.228975 0.134222 -0.331073 -0.048074 -0.584594 5
    3 -1.384011 -0.264878 0.411372 0.664069 -0.978798 -0.461036 -0.584594 6
    4 0.961576 -0.243630 -0.379014 0.558100 1.288240 -0.579025 -0.959946 5
    ... ... ... ... ... ... ... ... ...
    1594 0.403103 0.053829 -0.075020 -0.978459 0.899605 -0.461036 0.072271 5
    1595 0.123866 -0.541090 0.137777 -0.861893 1.353012 0.600867 0.729136 6
    1596 -0.099523 -0.243630 -0.196617 -0.533387 0.705287 0.541872 0.541460 6
    1597 0.654416 -0.264878 -0.075020 -0.676446 1.676875 0.305894 -0.209243 5
    1598 -1.216469 -0.434854 -0.135818 -0.665849 0.510970 0.010921 0.541460 6

    1599 rows × 8 columns

    In [312]:
    X = After_backward_elimination.drop(columns=['Signal_Strength']).values
Y = After_backward_elimination['Signal_Strength'].values
X_train, X_test, y_train, y_test = train_test_split(X, Y,test_size=0.33, random_state=42)
    In [313]:
    X.shape
    Out[313]:
    (1599, 7)
    In [314]:
    model1 = Sequential()
model1.add(Dense(10, input_dim=7, kernel_initializer='normal', activation='relu'))
model1.add(Dropout(0.25))
model1.add(Dense(10, activation='relu'))
model1.add(Dropout(0.25))
model1.add(Dense(1, activation='linear'))
model1.summary()

    Model: "sequential_19"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_58 (Dense)             (None, 10)                80        
_________________________________________________________________
dropout_32 (Dropout)         (None, 10)                0         
_________________________________________________________________
dense_59 (Dense)             (None, 10)                110       
_________________________________________________________________
dropout_33 (Dropout)         (None, 10)                0         
_________________________________________________________________
dense_60 (Dense)             (None, 1)                 11        
=================================================================
Total params: 201
Trainable params: 201
Non-trainable params: 0
_________________________________________________________________
    In [315]:
    model1.compile(loss='mse', optimizer='adam', metrics=['mse','mae'])
    In [316]:
    history = model1.fit(X_train, y_train, epochs=200, batch_size=10,  verbose=1, validation_split=0.2)

    Train on 856 samples, validate on 215 samples
Epoch 1/200
856/856 [==============================] - 1s 713us/sample - loss: 27.0769 - mean_squared_error: 27.0769 - mean_absolute_error: 5.1202 - val_loss: 21.0025 - val_mean_squared_error: 21.0025 - val_mean_absolute_error: 4.5084
Epoch 2/200
856/856 [==============================] - 0s 153us/sample - loss: 14.2819 - mean_squared_error: 14.2819 - mean_absolute_error: 3.5655 - val_loss: 6.6728 - val_mean_squared_error: 6.6728 - val_mean_absolute_error: 2.4240
Epoch 3/200
856/856 [==============================] - 0s 157us/sample - loss: 5.8012 - mean_squared_error: 5.8012 - mean_absolute_error: 2.0230 - val_loss: 1.7196 - val_mean_squared_error: 1.7196 - val_mean_absolute_error: 1.1247
Epoch 4/200
856/856 [==============================] - 0s 166us/sample - loss: 3.6995 - mean_squared_error: 3.6995 - mean_absolute_error: 1.5689 - val_loss: 0.9809 - val_mean_squared_error: 0.9809 - val_mean_absolute_error: 0.7907
Epoch 5/200
856/856 [==============================] - 0s 158us/sample - loss: 3.4464 - mean_squared_error: 3.4464 - mean_absolute_error: 1.4866 - val_loss: 0.8083 - val_mean_squared_error: 0.8083 - val_mean_absolute_error: 0.7176
Epoch 6/200
856/856 [==============================] - 0s 158us/sample - loss: 3.4701 - mean_squared_error: 3.4701 - mean_absolute_error: 1.4903 - val_loss: 0.7205 - val_mean_squared_error: 0.7205 - val_mean_absolute_error: 0.6643
Epoch 7/200
856/856 [==============================] - 0s 159us/sample - loss: 3.0271 - mean_squared_error: 3.0271 - mean_absolute_error: 1.3829 - val_loss: 0.6496 - val_mean_squared_error: 0.6496 - val_mean_absolute_error: 0.6268
Epoch 8/200
856/856 [==============================] - 0s 157us/sample - loss: 2.6860 - mean_squared_error: 2.6860 - mean_absolute_error: 1.3193 - val_loss: 0.5993 - val_mean_squared_error: 0.5993 - val_mean_absolute_error: 0.6042
Epoch 9/200
856/856 [==============================] - 0s 164us/sample - loss: 2.5391 - mean_squared_error: 2.5391 - mean_absolute_error: 1.2836 - val_loss: 0.4838 - val_mean_squared_error: 0.4838 - val_mean_absolute_error: 0.5333
Epoch 10/200
856/856 [==============================] - 0s 158us/sample - loss: 2.5447 - mean_squared_error: 2.5447 - mean_absolute_error: 1.2809 - val_loss: 0.5368 - val_mean_squared_error: 0.5368 - val_mean_absolute_error: 0.5804
Epoch 11/200
856/856 [==============================] - 0s 162us/sample - loss: 2.5914 - mean_squared_error: 2.5914 - mean_absolute_error: 1.2881 - val_loss: 0.4532 - val_mean_squared_error: 0.4532 - val_mean_absolute_error: 0.5244
Epoch 12/200
856/856 [==============================] - 0s 165us/sample - loss: 2.5336 - mean_squared_error: 2.5336 - mean_absolute_error: 1.2786 - val_loss: 0.4985 - val_mean_squared_error: 0.4985 - val_mean_absolute_error: 0.5529
Epoch 13/200
856/856 [==============================] - 0s 151us/sample - loss: 2.4436 - mean_squared_error: 2.4436 - mean_absolute_error: 1.2506 - val_loss: 0.5087 - val_mean_squared_error: 0.5087 - val_mean_absolute_error: 0.5666
Epoch 14/200
856/856 [==============================] - 0s 158us/sample - loss: 2.2224 - mean_squared_error: 2.2224 - mean_absolute_error: 1.1857 - val_loss: 0.4428 - val_mean_squared_error: 0.4428 - val_mean_absolute_error: 0.5094
Epoch 15/200
856/856 [==============================] - 0s 160us/sample - loss: 2.4097 - mean_squared_error: 2.4097 - mean_absolute_error: 1.2289 - val_loss: 0.4791 - val_mean_squared_error: 0.4791 - val_mean_absolute_error: 0.5405
Epoch 16/200
856/856 [==============================] - 0s 159us/sample - loss: 2.1911 - mean_squared_error: 2.1911 - mean_absolute_error: 1.1879 - val_loss: 0.5286 - val_mean_squared_error: 0.5286 - val_mean_absolute_error: 0.5771
Epoch 17/200
856/856 [==============================] - 0s 148us/sample - loss: 2.1595 - mean_squared_error: 2.1595 - mean_absolute_error: 1.1797 - val_loss: 0.5149 - val_mean_squared_error: 0.5149 - val_mean_absolute_error: 0.5585
Epoch 18/200
856/856 [==============================] - 0s 155us/sample - loss: 2.1089 - mean_squared_error: 2.1089 - mean_absolute_error: 1.1614 - val_loss: 0.4248 - val_mean_squared_error: 0.4248 - val_mean_absolute_error: 0.4939
Epoch 19/200
856/856 [==============================] - 0s 175us/sample - loss: 2.1095 - mean_squared_error: 2.1095 - mean_absolute_error: 1.1711 - val_loss: 0.4763 - val_mean_squared_error: 0.4763 - val_mean_absolute_error: 0.5404
Epoch 20/200
856/856 [==============================] - 0s 169us/sample - loss: 2.0930 - mean_squared_error: 2.0930 - mean_absolute_error: 1.1383 - val_loss: 0.4934 - val_mean_squared_error: 0.4934 - val_mean_absolute_error: 0.5390
Epoch 21/200
856/856 [==============================] - 0s 169us/sample - loss: 2.0752 - mean_squared_error: 2.0752 - mean_absolute_error: 1.1282 - val_loss: 0.4523 - val_mean_squared_error: 0.4523 - val_mean_absolute_error: 0.5144
Epoch 22/200
856/856 [==============================] - 0s 222us/sample - loss: 1.9387 - mean_squared_error: 1.9387 - mean_absolute_error: 1.1009 - val_loss: 0.4624 - val_mean_squared_error: 0.4624 - val_mean_absolute_error: 0.5183
Epoch 23/200
856/856 [==============================] - 0s 181us/sample - loss: 1.8747 - mean_squared_error: 1.8747 - mean_absolute_error: 1.0867 - val_loss: 0.3506 - val_mean_squared_error: 0.3506 - val_mean_absolute_error: 0.4441
Epoch 24/200
856/856 [==============================] - 0s 155us/sample - loss: 1.7527 - mean_squared_error: 1.7527 - mean_absolute_error: 1.0586 - val_loss: 0.4554 - val_mean_squared_error: 0.4554 - val_mean_absolute_error: 0.5172
Epoch 25/200
856/856 [==============================] - 0s 162us/sample - loss: 1.7567 - mean_squared_error: 1.7567 - mean_absolute_error: 1.0541 - val_loss: 0.4554 - val_mean_squared_error: 0.4554 - val_mean_absolute_error: 0.5285
Epoch 26/200
856/856 [==============================] - 0s 174us/sample - loss: 1.6403 - mean_squared_error: 1.6403 - mean_absolute_error: 1.0020 - val_loss: 0.3817 - val_mean_squared_error: 0.3817 - val_mean_absolute_error: 0.4671
Epoch 27/200
856/856 [==============================] - 0s 162us/sample - loss: 1.8577 - mean_squared_error: 1.8577 - mean_absolute_error: 1.0681 - val_loss: 0.4780 - val_mean_squared_error: 0.4780 - val_mean_absolute_error: 0.5300
Epoch 28/200
856/856 [==============================] - 0s 153us/sample - loss: 1.6038 - mean_squared_error: 1.6038 - mean_absolute_error: 0.9878 - val_loss: 0.4114 - val_mean_squared_error: 0.4114 - val_mean_absolute_error: 0.4873
Epoch 29/200
856/856 [==============================] - 0s 192us/sample - loss: 1.6754 - mean_squared_error: 1.6754 - mean_absolute_error: 1.0245 - val_loss: 0.3951 - val_mean_squared_error: 0.3951 - val_mean_absolute_error: 0.4748
Epoch 30/200
856/856 [==============================] - 0s 166us/sample - loss: 1.6808 - mean_squared_error: 1.6808 - mean_absolute_error: 1.0188 - val_loss: 0.4320 - val_mean_squared_error: 0.4320 - val_mean_absolute_error: 0.4989
Epoch 31/200
856/856 [==============================] - 0s 165us/sample - loss: 1.6089 - mean_squared_error: 1.6089 - mean_absolute_error: 0.9751 - val_loss: 0.4059 - val_mean_squared_error: 0.4059 - val_mean_absolute_error: 0.4845
Epoch 32/200
856/856 [==============================] - 0s 183us/sample - loss: 1.6332 - mean_squared_error: 1.6332 - mean_absolute_error: 1.0078 - val_loss: 0.3744 - val_mean_squared_error: 0.3744 - val_mean_absolute_error: 0.4632
Epoch 33/200
856/856 [==============================] - 0s 155us/sample - loss: 1.5424 - mean_squared_error: 1.5424 - mean_absolute_error: 0.9934 - val_loss: 0.3782 - val_mean_squared_error: 0.3782 - val_mean_absolute_error: 0.4674
Epoch 34/200
856/856 [==============================] - 0s 181us/sample - loss: 1.6157 - mean_squared_error: 1.6157 - mean_absolute_error: 0.9932 - val_loss: 0.3925 - val_mean_squared_error: 0.3925 - val_mean_absolute_error: 0.4741
Epoch 35/200
856/856 [==============================] - 0s 159us/sample - loss: 1.5511 - mean_squared_error: 1.5511 - mean_absolute_error: 0.9932 - val_loss: 0.3517 - val_mean_squared_error: 0.3517 - val_mean_absolute_error: 0.4419
Epoch 36/200
856/856 [==============================] - 0s 153us/sample - loss: 1.5220 - mean_squared_error: 1.5220 - mean_absolute_error: 0.9721 - val_loss: 0.3852 - val_mean_squared_error: 0.3852 - val_mean_absolute_error: 0.4673
Epoch 37/200
856/856 [==============================] - 0s 137us/sample - loss: 1.5409 - mean_squared_error: 1.5409 - mean_absolute_error: 0.9872 - val_loss: 0.3771 - val_mean_squared_error: 0.3771 - val_mean_absolute_error: 0.4640
Epoch 38/200
856/856 [==============================] - 0s 139us/sample - loss: 1.5172 - mean_squared_error: 1.5172 - mean_absolute_error: 0.9738 - val_loss: 0.3866 - val_mean_squared_error: 0.3866 - val_mean_absolute_error: 0.4750
Epoch 39/200
856/856 [==============================] - 0s 141us/sample - loss: 1.4369 - mean_squared_error: 1.4369 - mean_absolute_error: 0.9417 - val_loss: 0.3820 - val_mean_squared_error: 0.3820 - val_mean_absolute_error: 0.4695
Epoch 40/200
856/856 [==============================] - 0s 146us/sample - loss: 1.4758 - mean_squared_error: 1.4758 - mean_absolute_error: 0.9479 - val_loss: 0.3766 - val_mean_squared_error: 0.3766 - val_mean_absolute_error: 0.4643
Epoch 41/200
856/856 [==============================] - 0s 174us/sample - loss: 1.3019 - mean_squared_error: 1.3019 - mean_absolute_error: 0.8998 - val_loss: 0.3973 - val_mean_squared_error: 0.3973 - val_mean_absolute_error: 0.4845
Epoch 42/200
856/856 [==============================] - 0s 172us/sample - loss: 1.3523 - mean_squared_error: 1.3523 - mean_absolute_error: 0.9095 - val_loss: 0.3887 - val_mean_squared_error: 0.3887 - val_mean_absolute_error: 0.4737
Epoch 43/200
856/856 [==============================] - 0s 144us/sample - loss: 1.4531 - mean_squared_error: 1.4531 - mean_absolute_error: 0.9650 - val_loss: 0.3811 - val_mean_squared_error: 0.3811 - val_mean_absolute_error: 0.4682
Epoch 44/200
856/856 [==============================] - 0s 180us/sample - loss: 1.4338 - mean_squared_error: 1.4338 - mean_absolute_error: 0.9579 - val_loss: 0.3536 - val_mean_squared_error: 0.3536 - val_mean_absolute_error: 0.4512
Epoch 45/200
856/856 [==============================] - 0s 188us/sample - loss: 1.3920 - mean_squared_error: 1.3920 - mean_absolute_error: 0.9272 - val_loss: 0.3738 - val_mean_squared_error: 0.3738 - val_mean_absolute_error: 0.4605
Epoch 46/200
856/856 [==============================] - 0s 154us/sample - loss: 1.4071 - mean_squared_error: 1.4071 - mean_absolute_error: 0.9307 - val_loss: 0.3715 - val_mean_squared_error: 0.3715 - val_mean_absolute_error: 0.4594
Epoch 47/200
856/856 [==============================] - 0s 193us/sample - loss: 1.2651 - mean_squared_error: 1.2651 - mean_absolute_error: 0.8719 - val_loss: 0.3524 - val_mean_squared_error: 0.3524 - val_mean_absolute_error: 0.4471
Epoch 48/200
856/856 [==============================] - 0s 138us/sample - loss: 1.3630 - mean_squared_error: 1.3630 - mean_absolute_error: 0.9229 - val_loss: 0.3631 - val_mean_squared_error: 0.3631 - val_mean_absolute_error: 0.4487
Epoch 49/200
856/856 [==============================] - 0s 150us/sample - loss: 1.2453 - mean_squared_error: 1.2453 - mean_absolute_error: 0.8896 - val_loss: 0.3463 - val_mean_squared_error: 0.3463 - val_mean_absolute_error: 0.4383
Epoch 50/200
856/856 [==============================] - 0s 151us/sample - loss: 1.3090 - mean_squared_error: 1.3090 - mean_absolute_error: 0.9063 - val_loss: 0.3582 - val_mean_squared_error: 0.3582 - val_mean_absolute_error: 0.4460
Epoch 51/200
856/856 [==============================] - 0s 148us/sample - loss: 1.1917 - mean_squared_error: 1.1917 - mean_absolute_error: 0.8614 - val_loss: 0.3440 - val_mean_squared_error: 0.3440 - val_mean_absolute_error: 0.4422
Epoch 52/200
856/856 [==============================] - 0s 141us/sample - loss: 1.2214 - mean_squared_error: 1.2214 - mean_absolute_error: 0.8671 - val_loss: 0.3467 - val_mean_squared_error: 0.3467 - val_mean_absolute_error: 0.4439
Epoch 53/200
856/856 [==============================] - 0s 148us/sample - loss: 1.2467 - mean_squared_error: 1.2467 - mean_absolute_error: 0.8850 - val_loss: 0.3681 - val_mean_squared_error: 0.3681 - val_mean_absolute_error: 0.4513
Epoch 54/200
856/856 [==============================] - 0s 148us/sample - loss: 1.2884 - mean_squared_error: 1.2884 - mean_absolute_error: 0.8921 - val_loss: 0.4086 - val_mean_squared_error: 0.4086 - val_mean_absolute_error: 0.4812
Epoch 55/200
856/856 [==============================] - 0s 140us/sample - loss: 1.2801 - mean_squared_error: 1.2801 - mean_absolute_error: 0.8881 - val_loss: 0.3907 - val_mean_squared_error: 0.3907 - val_mean_absolute_error: 0.4760
Epoch 56/200
856/856 [==============================] - 0s 147us/sample - loss: 1.1858 - mean_squared_error: 1.1858 - mean_absolute_error: 0.8667 - val_loss: 0.3469 - val_mean_squared_error: 0.3469 - val_mean_absolute_error: 0.4459
Epoch 57/200
856/856 [==============================] - 0s 151us/sample - loss: 1.1542 - mean_squared_error: 1.1542 - mean_absolute_error: 0.8430 - val_loss: 0.3541 - val_mean_squared_error: 0.3541 - val_mean_absolute_error: 0.4492
Epoch 58/200
856/856 [==============================] - 0s 162us/sample - loss: 1.1515 - mean_squared_error: 1.1515 - mean_absolute_error: 0.8466 - val_loss: 0.3560 - val_mean_squared_error: 0.3560 - val_mean_absolute_error: 0.4522
Epoch 59/200
856/856 [==============================] - 0s 166us/sample - loss: 1.1508 - mean_squared_error: 1.1508 - mean_absolute_error: 0.8495 - val_loss: 0.4014 - val_mean_squared_error: 0.4014 - val_mean_absolute_error: 0.4813
Epoch 60/200
856/856 [==============================] - 0s 132us/sample - loss: 1.1728 - mean_squared_error: 1.1728 - mean_absolute_error: 0.8437 - val_loss: 0.3930 - val_mean_squared_error: 0.3930 - val_mean_absolute_error: 0.4731
Epoch 61/200
856/856 [==============================] - 0s 138us/sample - loss: 1.1526 - mean_squared_error: 1.1526 - mean_absolute_error: 0.8308 - val_loss: 0.3308 - val_mean_squared_error: 0.3308 - val_mean_absolute_error: 0.4403
Epoch 62/200
856/856 [==============================] - 0s 150us/sample - loss: 1.0861 - mean_squared_error: 1.0861 - mean_absolute_error: 0.8194 - val_loss: 0.3616 - val_mean_squared_error: 0.3616 - val_mean_absolute_error: 0.4590
Epoch 63/200
856/856 [==============================] - 0s 147us/sample - loss: 1.1102 - mean_squared_error: 1.1102 - mean_absolute_error: 0.8256 - val_loss: 0.3528 - val_mean_squared_error: 0.3528 - val_mean_absolute_error: 0.4587
Epoch 64/200
856/856 [==============================] - 0s 141us/sample - loss: 1.0456 - mean_squared_error: 1.0456 - mean_absolute_error: 0.8000 - val_loss: 0.3362 - val_mean_squared_error: 0.3362 - val_mean_absolute_error: 0.4472
Epoch 65/200
856/856 [==============================] - 0s 148us/sample - loss: 1.1004 - mean_squared_error: 1.1004 - mean_absolute_error: 0.8299 - val_loss: 0.3534 - val_mean_squared_error: 0.3534 - val_mean_absolute_error: 0.4521
Epoch 66/200
856/856 [==============================] - 0s 151us/sample - loss: 1.0562 - mean_squared_error: 1.0562 - mean_absolute_error: 0.7887 - val_loss: 0.3667 - val_mean_squared_error: 0.3667 - val_mean_absolute_error: 0.4633
Epoch 67/200
856/856 [==============================] - 0s 145us/sample - loss: 1.0206 - mean_squared_error: 1.0206 - mean_absolute_error: 0.7946 - val_loss: 0.3458 - val_mean_squared_error: 0.3458 - val_mean_absolute_error: 0.4516
Epoch 68/200
856/856 [==============================] - 0s 148us/sample - loss: 1.0377 - mean_squared_error: 1.0377 - mean_absolute_error: 0.8007 - val_loss: 0.3460 - val_mean_squared_error: 0.3460 - val_mean_absolute_error: 0.4463
Epoch 69/200
856/856 [==============================] - 0s 151us/sample - loss: 1.0853 - mean_squared_error: 1.0853 - mean_absolute_error: 0.8061 - val_loss: 0.3234 - val_mean_squared_error: 0.3234 - val_mean_absolute_error: 0.4358
Epoch 70/200
856/856 [==============================] - 0s 144us/sample - loss: 0.9998 - mean_squared_error: 0.9998 - mean_absolute_error: 0.7890 - val_loss: 0.3292 - val_mean_squared_error: 0.3292 - val_mean_absolute_error: 0.4418
Epoch 71/200
856/856 [==============================] - 0s 155us/sample - loss: 1.0049 - mean_squared_error: 1.0049 - mean_absolute_error: 0.7845 - val_loss: 0.3320 - val_mean_squared_error: 0.3320 - val_mean_absolute_error: 0.4424
Epoch 72/200
856/856 [==============================] - 0s 145us/sample - loss: 0.9582 - mean_squared_error: 0.9582 - mean_absolute_error: 0.7731 - val_loss: 0.3633 - val_mean_squared_error: 0.3633 - val_mean_absolute_error: 0.4601
Epoch 73/200
856/856 [==============================] - 0s 148us/sample - loss: 1.0058 - mean_squared_error: 1.0058 - mean_absolute_error: 0.7918 - val_loss: 0.3568 - val_mean_squared_error: 0.3568 - val_mean_absolute_error: 0.4551
Epoch 74/200
856/856 [==============================] - 0s 153us/sample - loss: 0.9598 - mean_squared_error: 0.9598 - mean_absolute_error: 0.7609 - val_loss: 0.3578 - val_mean_squared_error: 0.3578 - val_mean_absolute_error: 0.4598
Epoch 75/200
856/856 [==============================] - 0s 153us/sample - loss: 0.9917 - mean_squared_error: 0.9917 - mean_absolute_error: 0.7749 - val_loss: 0.3425 - val_mean_squared_error: 0.3425 - val_mean_absolute_error: 0.4550
Epoch 76/200
856/856 [==============================] - 0s 151us/sample - loss: 1.0454 - mean_squared_error: 1.0454 - mean_absolute_error: 0.8074 - val_loss: 0.3489 - val_mean_squared_error: 0.3489 - val_mean_absolute_error: 0.4545
Epoch 77/200
856/856 [==============================] - 0s 150us/sample - loss: 0.8903 - mean_squared_error: 0.8903 - mean_absolute_error: 0.7416 - val_loss: 0.3349 - val_mean_squared_error: 0.3349 - val_mean_absolute_error: 0.4416
Epoch 78/200
856/856 [==============================] - 0s 145us/sample - loss: 1.0474 - mean_squared_error: 1.0474 - mean_absolute_error: 0.8027 - val_loss: 0.3336 - val_mean_squared_error: 0.3336 - val_mean_absolute_error: 0.4393
Epoch 79/200
856/856 [==============================] - 0s 148us/sample - loss: 0.9318 - mean_squared_error: 0.9318 - mean_absolute_error: 0.7625 - val_loss: 0.3563 - val_mean_squared_error: 0.3563 - val_mean_absolute_error: 0.4522
Epoch 80/200
856/856 [==============================] - 0s 150us/sample - loss: 0.9156 - mean_squared_error: 0.9156 - mean_absolute_error: 0.7320 - val_loss: 0.3394 - val_mean_squared_error: 0.3394 - val_mean_absolute_error: 0.4470
Epoch 81/200
856/856 [==============================] - 0s 147us/sample - loss: 0.9023 - mean_squared_error: 0.9023 - mean_absolute_error: 0.7364 - val_loss: 0.3454 - val_mean_squared_error: 0.3454 - val_mean_absolute_error: 0.4509
Epoch 82/200
856/856 [==============================] - 0s 148us/sample - loss: 0.9380 - mean_squared_error: 0.9380 - mean_absolute_error: 0.7613 - val_loss: 0.3445 - val_mean_squared_error: 0.3445 - val_mean_absolute_error: 0.4529
Epoch 83/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8765 - mean_squared_error: 0.8765 - mean_absolute_error: 0.7236 - val_loss: 0.3277 - val_mean_squared_error: 0.3277 - val_mean_absolute_error: 0.4413
Epoch 84/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8396 - mean_squared_error: 0.8396 - mean_absolute_error: 0.7351 - val_loss: 0.3638 - val_mean_squared_error: 0.3638 - val_mean_absolute_error: 0.4636
Epoch 85/200
856/856 [==============================] - 0s 147us/sample - loss: 0.9005 - mean_squared_error: 0.9005 - mean_absolute_error: 0.7382 - val_loss: 0.3437 - val_mean_squared_error: 0.3437 - val_mean_absolute_error: 0.4445
Epoch 86/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8756 - mean_squared_error: 0.8756 - mean_absolute_error: 0.7336 - val_loss: 0.3269 - val_mean_squared_error: 0.3269 - val_mean_absolute_error: 0.4351
Epoch 87/200
856/856 [==============================] - 0s 146us/sample - loss: 0.8850 - mean_squared_error: 0.8850 - mean_absolute_error: 0.7356 - val_loss: 0.3281 - val_mean_squared_error: 0.3281 - val_mean_absolute_error: 0.4428
Epoch 88/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8428 - mean_squared_error: 0.8428 - mean_absolute_error: 0.7247 - val_loss: 0.3386 - val_mean_squared_error: 0.3386 - val_mean_absolute_error: 0.4488
Epoch 89/200
856/856 [==============================] - 0s 148us/sample - loss: 0.8499 - mean_squared_error: 0.8499 - mean_absolute_error: 0.7139 - val_loss: 0.3544 - val_mean_squared_error: 0.3544 - val_mean_absolute_error: 0.4540
Epoch 90/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8551 - mean_squared_error: 0.8551 - mean_absolute_error: 0.7257 - val_loss: 0.3364 - val_mean_squared_error: 0.3364 - val_mean_absolute_error: 0.4395
Epoch 91/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8303 - mean_squared_error: 0.8303 - mean_absolute_error: 0.7135 - val_loss: 0.3316 - val_mean_squared_error: 0.3316 - val_mean_absolute_error: 0.4395
Epoch 92/200
856/856 [==============================] - 0s 141us/sample - loss: 0.8816 - mean_squared_error: 0.8816 - mean_absolute_error: 0.7322 - val_loss: 0.3319 - val_mean_squared_error: 0.3319 - val_mean_absolute_error: 0.4416
Epoch 93/200
856/856 [==============================] - 0s 147us/sample - loss: 0.7980 - mean_squared_error: 0.7980 - mean_absolute_error: 0.7005 - val_loss: 0.3538 - val_mean_squared_error: 0.3538 - val_mean_absolute_error: 0.4545
Epoch 94/200
856/856 [==============================] - 0s 148us/sample - loss: 0.8809 - mean_squared_error: 0.8809 - mean_absolute_error: 0.7244 - val_loss: 0.3503 - val_mean_squared_error: 0.3503 - val_mean_absolute_error: 0.4528
Epoch 95/200
856/856 [==============================] - 0s 140us/sample - loss: 0.7647 - mean_squared_error: 0.7647 - mean_absolute_error: 0.6771 - val_loss: 0.3430 - val_mean_squared_error: 0.3430 - val_mean_absolute_error: 0.4490
Epoch 96/200
856/856 [==============================] - 0s 147us/sample - loss: 0.8204 - mean_squared_error: 0.8204 - mean_absolute_error: 0.7058 - val_loss: 0.3502 - val_mean_squared_error: 0.3502 - val_mean_absolute_error: 0.4555
Epoch 97/200
856/856 [==============================] - 0s 152us/sample - loss: 0.8228 - mean_squared_error: 0.8228 - mean_absolute_error: 0.7079 - val_loss: 0.3420 - val_mean_squared_error: 0.3420 - val_mean_absolute_error: 0.4528
Epoch 98/200
856/856 [==============================] - 0s 171us/sample - loss: 0.7235 - mean_squared_error: 0.7235 - mean_absolute_error: 0.6641 - val_loss: 0.3368 - val_mean_squared_error: 0.3368 - val_mean_absolute_error: 0.4521
Epoch 99/200
856/856 [==============================] - 0s 147us/sample - loss: 0.7540 - mean_squared_error: 0.7540 - mean_absolute_error: 0.6772 - val_loss: 0.3203 - val_mean_squared_error: 0.3203 - val_mean_absolute_error: 0.4335
Epoch 100/200
856/856 [==============================] - 0s 136us/sample - loss: 0.7806 - mean_squared_error: 0.7806 - mean_absolute_error: 0.6954 - val_loss: 0.3295 - val_mean_squared_error: 0.3295 - val_mean_absolute_error: 0.4444
Epoch 101/200
856/856 [==============================] - 0s 141us/sample - loss: 0.7549 - mean_squared_error: 0.7549 - mean_absolute_error: 0.6761 - val_loss: 0.3408 - val_mean_squared_error: 0.3408 - val_mean_absolute_error: 0.4498
Epoch 102/200
856/856 [==============================] - 0s 146us/sample - loss: 0.8110 - mean_squared_error: 0.8110 - mean_absolute_error: 0.6964 - val_loss: 0.3438 - val_mean_squared_error: 0.3438 - val_mean_absolute_error: 0.4509
Epoch 103/200
856/856 [==============================] - 0s 145us/sample - loss: 0.6770 - mean_squared_error: 0.6770 - mean_absolute_error: 0.6477 - val_loss: 0.3366 - val_mean_squared_error: 0.3366 - val_mean_absolute_error: 0.4512
Epoch 104/200
856/856 [==============================] - 0s 148us/sample - loss: 0.7376 - mean_squared_error: 0.7376 - mean_absolute_error: 0.6692 - val_loss: 0.3405 - val_mean_squared_error: 0.3405 - val_mean_absolute_error: 0.4473
Epoch 105/200
856/856 [==============================] - 0s 154us/sample - loss: 0.7675 - mean_squared_error: 0.7675 - mean_absolute_error: 0.6791 - val_loss: 0.3280 - val_mean_squared_error: 0.3280 - val_mean_absolute_error: 0.4446
Epoch 106/200
856/856 [==============================] - 0s 150us/sample - loss: 0.7151 - mean_squared_error: 0.7151 - mean_absolute_error: 0.6583 - val_loss: 0.3569 - val_mean_squared_error: 0.3569 - val_mean_absolute_error: 0.4670
Epoch 107/200
856/856 [==============================] - 0s 151us/sample - loss: 0.7248 - mean_squared_error: 0.7248 - mean_absolute_error: 0.6721 - val_loss: 0.3320 - val_mean_squared_error: 0.3320 - val_mean_absolute_error: 0.4479
Epoch 108/200
856/856 [==============================] - 0s 147us/sample - loss: 0.7621 - mean_squared_error: 0.7621 - mean_absolute_error: 0.6828 - val_loss: 0.3273 - val_mean_squared_error: 0.3273 - val_mean_absolute_error: 0.4449
Epoch 109/200
856/856 [==============================] - 0s 150us/sample - loss: 0.7629 - mean_squared_error: 0.7629 - mean_absolute_error: 0.6749 - val_loss: 0.3253 - val_mean_squared_error: 0.3253 - val_mean_absolute_error: 0.4452
Epoch 110/200
856/856 [==============================] - 0s 152us/sample - loss: 0.7198 - mean_squared_error: 0.7198 - mean_absolute_error: 0.6570 - val_loss: 0.3193 - val_mean_squared_error: 0.3193 - val_mean_absolute_error: 0.4373
Epoch 111/200
856/856 [==============================] - 0s 154us/sample - loss: 0.6877 - mean_squared_error: 0.6877 - mean_absolute_error: 0.6425 - val_loss: 0.3148 - val_mean_squared_error: 0.3148 - val_mean_absolute_error: 0.4314
Epoch 112/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6775 - mean_squared_error: 0.6775 - mean_absolute_error: 0.6426 - val_loss: 0.3536 - val_mean_squared_error: 0.3536 - val_mean_absolute_error: 0.4584
Epoch 113/200
856/856 [==============================] - 0s 150us/sample - loss: 0.7140 - mean_squared_error: 0.7140 - mean_absolute_error: 0.6614 - val_loss: 0.3358 - val_mean_squared_error: 0.3358 - val_mean_absolute_error: 0.4465
Epoch 114/200
856/856 [==============================] - 0s 151us/sample - loss: 0.6659 - mean_squared_error: 0.6659 - mean_absolute_error: 0.6387 - val_loss: 0.3237 - val_mean_squared_error: 0.3237 - val_mean_absolute_error: 0.4410
Epoch 115/200
856/856 [==============================] - 0s 143us/sample - loss: 0.6399 - mean_squared_error: 0.6399 - mean_absolute_error: 0.6194 - val_loss: 0.3329 - val_mean_squared_error: 0.3329 - val_mean_absolute_error: 0.4462
Epoch 116/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6403 - mean_squared_error: 0.6403 - mean_absolute_error: 0.6342 - val_loss: 0.3203 - val_mean_squared_error: 0.3203 - val_mean_absolute_error: 0.4395
Epoch 117/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6597 - mean_squared_error: 0.6597 - mean_absolute_error: 0.6380 - val_loss: 0.3219 - val_mean_squared_error: 0.3219 - val_mean_absolute_error: 0.4396
Epoch 118/200
856/856 [==============================] - 0s 141us/sample - loss: 0.6568 - mean_squared_error: 0.6568 - mean_absolute_error: 0.6446 - val_loss: 0.3294 - val_mean_squared_error: 0.3294 - val_mean_absolute_error: 0.4427
Epoch 119/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6384 - mean_squared_error: 0.6384 - mean_absolute_error: 0.6169 - val_loss: 0.3200 - val_mean_squared_error: 0.3200 - val_mean_absolute_error: 0.4380
Epoch 120/200
856/856 [==============================] - 0s 153us/sample - loss: 0.6309 - mean_squared_error: 0.6309 - mean_absolute_error: 0.6151 - val_loss: 0.3304 - val_mean_squared_error: 0.3304 - val_mean_absolute_error: 0.4451
Epoch 121/200
856/856 [==============================] - 0s 154us/sample - loss: 0.6597 - mean_squared_error: 0.6597 - mean_absolute_error: 0.6318 - val_loss: 0.3232 - val_mean_squared_error: 0.3232 - val_mean_absolute_error: 0.4385
Epoch 122/200
856/856 [==============================] - 0s 145us/sample - loss: 0.6317 - mean_squared_error: 0.6317 - mean_absolute_error: 0.6208 - val_loss: 0.3242 - val_mean_squared_error: 0.3242 - val_mean_absolute_error: 0.4373
Epoch 123/200
856/856 [==============================] - 0s 138us/sample - loss: 0.6360 - mean_squared_error: 0.6360 - mean_absolute_error: 0.6203 - val_loss: 0.3250 - val_mean_squared_error: 0.3250 - val_mean_absolute_error: 0.4441
Epoch 124/200
856/856 [==============================] - 0s 154us/sample - loss: 0.6326 - mean_squared_error: 0.6326 - mean_absolute_error: 0.6236 - val_loss: 0.3130 - val_mean_squared_error: 0.3130 - val_mean_absolute_error: 0.4353
Epoch 125/200
856/856 [==============================] - 0s 138us/sample - loss: 0.6136 - mean_squared_error: 0.6136 - mean_absolute_error: 0.6066 - val_loss: 0.3229 - val_mean_squared_error: 0.3229 - val_mean_absolute_error: 0.4435
Epoch 126/200
856/856 [==============================] - 0s 140us/sample - loss: 0.6667 - mean_squared_error: 0.6667 - mean_absolute_error: 0.6395 - val_loss: 0.3245 - val_mean_squared_error: 0.3245 - val_mean_absolute_error: 0.4450
Epoch 127/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5841 - mean_squared_error: 0.5841 - mean_absolute_error: 0.6057 - val_loss: 0.3191 - val_mean_squared_error: 0.3191 - val_mean_absolute_error: 0.4400
Epoch 128/200
856/856 [==============================] - 0s 146us/sample - loss: 0.6070 - mean_squared_error: 0.6070 - mean_absolute_error: 0.6041 - val_loss: 0.3261 - val_mean_squared_error: 0.3261 - val_mean_absolute_error: 0.4458
Epoch 129/200
856/856 [==============================] - 0s 140us/sample - loss: 0.5996 - mean_squared_error: 0.5996 - mean_absolute_error: 0.5942 - val_loss: 0.3231 - val_mean_squared_error: 0.3231 - val_mean_absolute_error: 0.4389
Epoch 130/200
856/856 [==============================] - 0s 140us/sample - loss: 0.6061 - mean_squared_error: 0.6061 - mean_absolute_error: 0.6027 - val_loss: 0.3209 - val_mean_squared_error: 0.3209 - val_mean_absolute_error: 0.4393
Epoch 131/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5877 - mean_squared_error: 0.5877 - mean_absolute_error: 0.5821 - val_loss: 0.3242 - val_mean_squared_error: 0.3242 - val_mean_absolute_error: 0.4436
Epoch 132/200
856/856 [==============================] - 0s 141us/sample - loss: 0.5463 - mean_squared_error: 0.5463 - mean_absolute_error: 0.5751 - val_loss: 0.3211 - val_mean_squared_error: 0.3211 - val_mean_absolute_error: 0.4404
Epoch 133/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5881 - mean_squared_error: 0.5881 - mean_absolute_error: 0.6052 - val_loss: 0.3242 - val_mean_squared_error: 0.3242 - val_mean_absolute_error: 0.4443
Epoch 134/200
856/856 [==============================] - 0s 152us/sample - loss: 0.6268 - mean_squared_error: 0.6268 - mean_absolute_error: 0.6166 - val_loss: 0.3271 - val_mean_squared_error: 0.3271 - val_mean_absolute_error: 0.4474
Epoch 135/200
856/856 [==============================] - 0s 146us/sample - loss: 0.6066 - mean_squared_error: 0.6066 - mean_absolute_error: 0.6071 - val_loss: 0.3262 - val_mean_squared_error: 0.3262 - val_mean_absolute_error: 0.4478
Epoch 136/200
856/856 [==============================] - 0s 153us/sample - loss: 0.5979 - mean_squared_error: 0.5979 - mean_absolute_error: 0.5967 - val_loss: 0.3221 - val_mean_squared_error: 0.3221 - val_mean_absolute_error: 0.4422
Epoch 137/200
856/856 [==============================] - 0s 147us/sample - loss: 0.6071 - mean_squared_error: 0.6071 - mean_absolute_error: 0.6031 - val_loss: 0.3174 - val_mean_squared_error: 0.3174 - val_mean_absolute_error: 0.4375
Epoch 138/200
856/856 [==============================] - 0s 186us/sample - loss: 0.5618 - mean_squared_error: 0.5618 - mean_absolute_error: 0.5875 - val_loss: 0.3279 - val_mean_squared_error: 0.3279 - val_mean_absolute_error: 0.4482
Epoch 139/200
856/856 [==============================] - 0s 155us/sample - loss: 0.5762 - mean_squared_error: 0.5762 - mean_absolute_error: 0.5863 - val_loss: 0.3226 - val_mean_squared_error: 0.3226 - val_mean_absolute_error: 0.4454
Epoch 140/200
856/856 [==============================] - 0s 158us/sample - loss: 0.5916 - mean_squared_error: 0.5916 - mean_absolute_error: 0.5942 - val_loss: 0.3268 - val_mean_squared_error: 0.3268 - val_mean_absolute_error: 0.4466
Epoch 141/200
856/856 [==============================] - 0s 164us/sample - loss: 0.5680 - mean_squared_error: 0.5680 - mean_absolute_error: 0.5822 - val_loss: 0.3219 - val_mean_squared_error: 0.3219 - val_mean_absolute_error: 0.4434
Epoch 142/200
856/856 [==============================] - 0s 157us/sample - loss: 0.5853 - mean_squared_error: 0.5853 - mean_absolute_error: 0.5854 - val_loss: 0.3294 - val_mean_squared_error: 0.3294 - val_mean_absolute_error: 0.4479
Epoch 143/200
856/856 [==============================] - 0s 165us/sample - loss: 0.5306 - mean_squared_error: 0.5306 - mean_absolute_error: 0.5737 - val_loss: 0.3271 - val_mean_squared_error: 0.3271 - val_mean_absolute_error: 0.4475
Epoch 144/200
856/856 [==============================] - 0s 133us/sample - loss: 0.5473 - mean_squared_error: 0.5473 - mean_absolute_error: 0.5699 - val_loss: 0.3250 - val_mean_squared_error: 0.3250 - val_mean_absolute_error: 0.4457
Epoch 145/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5292 - mean_squared_error: 0.5292 - mean_absolute_error: 0.5614 - val_loss: 0.3316 - val_mean_squared_error: 0.3316 - val_mean_absolute_error: 0.4503
Epoch 146/200
856/856 [==============================] - 0s 143us/sample - loss: 0.5552 - mean_squared_error: 0.5552 - mean_absolute_error: 0.5764 - val_loss: 0.3223 - val_mean_squared_error: 0.3223 - val_mean_absolute_error: 0.4435
Epoch 147/200
856/856 [==============================] - 0s 140us/sample - loss: 0.5457 - mean_squared_error: 0.5457 - mean_absolute_error: 0.5740 - val_loss: 0.3263 - val_mean_squared_error: 0.3263 - val_mean_absolute_error: 0.4477
Epoch 148/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5829 - mean_squared_error: 0.5829 - mean_absolute_error: 0.5819 - val_loss: 0.3327 - val_mean_squared_error: 0.3327 - val_mean_absolute_error: 0.4553
Epoch 149/200
856/856 [==============================] - 0s 144us/sample - loss: 0.5479 - mean_squared_error: 0.5479 - mean_absolute_error: 0.5739 - val_loss: 0.3296 - val_mean_squared_error: 0.3296 - val_mean_absolute_error: 0.4531
Epoch 150/200
856/856 [==============================] - 0s 143us/sample - loss: 0.5537 - mean_squared_error: 0.5537 - mean_absolute_error: 0.5743 - val_loss: 0.3269 - val_mean_squared_error: 0.3269 - val_mean_absolute_error: 0.4518
Epoch 151/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5512 - mean_squared_error: 0.5512 - mean_absolute_error: 0.5748 - val_loss: 0.3341 - val_mean_squared_error: 0.3341 - val_mean_absolute_error: 0.4557
Epoch 152/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5338 - mean_squared_error: 0.5338 - mean_absolute_error: 0.5586 - val_loss: 0.3259 - val_mean_squared_error: 0.3259 - val_mean_absolute_error: 0.4489
Epoch 153/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5383 - mean_squared_error: 0.5383 - mean_absolute_error: 0.5605 - val_loss: 0.3311 - val_mean_squared_error: 0.3311 - val_mean_absolute_error: 0.4509
Epoch 154/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5283 - mean_squared_error: 0.5283 - mean_absolute_error: 0.5565 - val_loss: 0.3250 - val_mean_squared_error: 0.3250 - val_mean_absolute_error: 0.4469
Epoch 155/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5281 - mean_squared_error: 0.5281 - mean_absolute_error: 0.5580 - val_loss: 0.3279 - val_mean_squared_error: 0.3279 - val_mean_absolute_error: 0.4504
Epoch 156/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5377 - mean_squared_error: 0.5377 - mean_absolute_error: 0.5606 - val_loss: 0.3364 - val_mean_squared_error: 0.3364 - val_mean_absolute_error: 0.4554
Epoch 157/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5383 - mean_squared_error: 0.5383 - mean_absolute_error: 0.5667 - val_loss: 0.3285 - val_mean_squared_error: 0.3285 - val_mean_absolute_error: 0.4505
Epoch 158/200
856/856 [==============================] - 0s 140us/sample - loss: 0.5124 - mean_squared_error: 0.5124 - mean_absolute_error: 0.5587 - val_loss: 0.3320 - val_mean_squared_error: 0.3320 - val_mean_absolute_error: 0.4530
Epoch 159/200
856/856 [==============================] - 0s 152us/sample - loss: 0.5204 - mean_squared_error: 0.5204 - mean_absolute_error: 0.5527 - val_loss: 0.3278 - val_mean_squared_error: 0.3278 - val_mean_absolute_error: 0.4491
Epoch 160/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5298 - mean_squared_error: 0.5298 - mean_absolute_error: 0.5601 - val_loss: 0.3388 - val_mean_squared_error: 0.3388 - val_mean_absolute_error: 0.4555
Epoch 161/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5316 - mean_squared_error: 0.5316 - mean_absolute_error: 0.5638 - val_loss: 0.3312 - val_mean_squared_error: 0.3312 - val_mean_absolute_error: 0.4535
Epoch 162/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5027 - mean_squared_error: 0.5027 - mean_absolute_error: 0.5552 - val_loss: 0.3309 - val_mean_squared_error: 0.3309 - val_mean_absolute_error: 0.4560
Epoch 163/200
856/856 [==============================] - 0s 146us/sample - loss: 0.5146 - mean_squared_error: 0.5146 - mean_absolute_error: 0.5552 - val_loss: 0.3356 - val_mean_squared_error: 0.3356 - val_mean_absolute_error: 0.4566
Epoch 164/200
856/856 [==============================] - 0s 148us/sample - loss: 0.4861 - mean_squared_error: 0.4861 - mean_absolute_error: 0.5410 - val_loss: 0.3261 - val_mean_squared_error: 0.3261 - val_mean_absolute_error: 0.4489
Epoch 165/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5042 - mean_squared_error: 0.5042 - mean_absolute_error: 0.5545 - val_loss: 0.3229 - val_mean_squared_error: 0.3229 - val_mean_absolute_error: 0.4456
Epoch 166/200
856/856 [==============================] - 0s 144us/sample - loss: 0.4974 - mean_squared_error: 0.4974 - mean_absolute_error: 0.5408 - val_loss: 0.3250 - val_mean_squared_error: 0.3250 - val_mean_absolute_error: 0.4456
Epoch 167/200
856/856 [==============================] - 0s 150us/sample - loss: 0.5154 - mean_squared_error: 0.5154 - mean_absolute_error: 0.5561 - val_loss: 0.3294 - val_mean_squared_error: 0.3294 - val_mean_absolute_error: 0.4509
Epoch 168/200
856/856 [==============================] - 0s 141us/sample - loss: 0.5094 - mean_squared_error: 0.5094 - mean_absolute_error: 0.5485 - val_loss: 0.3292 - val_mean_squared_error: 0.3292 - val_mean_absolute_error: 0.4492
Epoch 169/200
856/856 [==============================] - 0s 148us/sample - loss: 0.5087 - mean_squared_error: 0.5087 - mean_absolute_error: 0.5525 - val_loss: 0.3208 - val_mean_squared_error: 0.3208 - val_mean_absolute_error: 0.4444
Epoch 170/200
856/856 [==============================] - 0s 148us/sample - loss: 0.4773 - mean_squared_error: 0.4773 - mean_absolute_error: 0.5372 - val_loss: 0.3235 - val_mean_squared_error: 0.3235 - val_mean_absolute_error: 0.4466
Epoch 171/200
856/856 [==============================] - 0s 139us/sample - loss: 0.5081 - mean_squared_error: 0.5081 - mean_absolute_error: 0.5523 - val_loss: 0.3237 - val_mean_squared_error: 0.3237 - val_mean_absolute_error: 0.4473
Epoch 172/200
856/856 [==============================] - 0s 144us/sample - loss: 0.4962 - mean_squared_error: 0.4962 - mean_absolute_error: 0.5371 - val_loss: 0.3306 - val_mean_squared_error: 0.3306 - val_mean_absolute_error: 0.4527
Epoch 173/200
856/856 [==============================] - 0s 134us/sample - loss: 0.4816 - mean_squared_error: 0.4816 - mean_absolute_error: 0.5326 - val_loss: 0.3209 - val_mean_squared_error: 0.3209 - val_mean_absolute_error: 0.4463
Epoch 174/200
856/856 [==============================] - 0s 141us/sample - loss: 0.4807 - mean_squared_error: 0.4807 - mean_absolute_error: 0.5304 - val_loss: 0.3274 - val_mean_squared_error: 0.3274 - val_mean_absolute_error: 0.4519
Epoch 175/200
856/856 [==============================] - 0s 153us/sample - loss: 0.4835 - mean_squared_error: 0.4835 - mean_absolute_error: 0.5291 - val_loss: 0.3232 - val_mean_squared_error: 0.3232 - val_mean_absolute_error: 0.4480
Epoch 176/200
856/856 [==============================] - 0s 147us/sample - loss: 0.5052 - mean_squared_error: 0.5052 - mean_absolute_error: 0.5466 - val_loss: 0.3252 - val_mean_squared_error: 0.3252 - val_mean_absolute_error: 0.4509
Epoch 177/200
856/856 [==============================] - 0s 161us/sample - loss: 0.4732 - mean_squared_error: 0.4732 - mean_absolute_error: 0.5323 - val_loss: 0.3288 - val_mean_squared_error: 0.3288 - val_mean_absolute_error: 0.4488
Epoch 178/200
856/856 [==============================] - 0s 165us/sample - loss: 0.4931 - mean_squared_error: 0.4931 - mean_absolute_error: 0.5426 - val_loss: 0.3282 - val_mean_squared_error: 0.3282 - val_mean_absolute_error: 0.4478
Epoch 179/200
856/856 [==============================] - 0s 138us/sample - loss: 0.5424 - mean_squared_error: 0.5424 - mean_absolute_error: 0.5687 - val_loss: 0.3340 - val_mean_squared_error: 0.3340 - val_mean_absolute_error: 0.4526
Epoch 180/200
856/856 [==============================] - 0s 147us/sample - loss: 0.4497 - mean_squared_error: 0.4497 - mean_absolute_error: 0.5239 - val_loss: 0.3255 - val_mean_squared_error: 0.3255 - val_mean_absolute_error: 0.4479
Epoch 181/200
856/856 [==============================] - 0s 152us/sample - loss: 0.5013 - mean_squared_error: 0.5013 - mean_absolute_error: 0.5392 - val_loss: 0.3269 - val_mean_squared_error: 0.3269 - val_mean_absolute_error: 0.4474
Epoch 182/200
856/856 [==============================] - 0s 153us/sample - loss: 0.4942 - mean_squared_error: 0.4942 - mean_absolute_error: 0.5399 - val_loss: 0.3253 - val_mean_squared_error: 0.3253 - val_mean_absolute_error: 0.4442
Epoch 183/200
856/856 [==============================] - 0s 140us/sample - loss: 0.4810 - mean_squared_error: 0.4810 - mean_absolute_error: 0.5363 - val_loss: 0.3265 - val_mean_squared_error: 0.3265 - val_mean_absolute_error: 0.4427
Epoch 184/200
856/856 [==============================] - 0s 139us/sample - loss: 0.4749 - mean_squared_error: 0.4749 - mean_absolute_error: 0.5322 - val_loss: 0.3238 - val_mean_squared_error: 0.3238 - val_mean_absolute_error: 0.4431
Epoch 185/200
856/856 [==============================] - 0s 138us/sample - loss: 0.4823 - mean_squared_error: 0.4823 - mean_absolute_error: 0.5315 - val_loss: 0.3283 - val_mean_squared_error: 0.3283 - val_mean_absolute_error: 0.4480
Epoch 186/200
856/856 [==============================] - 0s 139us/sample - loss: 0.4795 - mean_squared_error: 0.4795 - mean_absolute_error: 0.5388 - val_loss: 0.3346 - val_mean_squared_error: 0.3346 - val_mean_absolute_error: 0.4528
Epoch 187/200
856/856 [==============================] - 0s 147us/sample - loss: 0.4788 - mean_squared_error: 0.4788 - mean_absolute_error: 0.5358 - val_loss: 0.3298 - val_mean_squared_error: 0.3298 - val_mean_absolute_error: 0.4484
Epoch 188/200
856/856 [==============================] - 0s 140us/sample - loss: 0.4865 - mean_squared_error: 0.4865 - mean_absolute_error: 0.5375 - val_loss: 0.3324 - val_mean_squared_error: 0.3324 - val_mean_absolute_error: 0.4488
Epoch 189/200
856/856 [==============================] - 0s 140us/sample - loss: 0.4803 - mean_squared_error: 0.4803 - mean_absolute_error: 0.5343 - val_loss: 0.3225 - val_mean_squared_error: 0.3225 - val_mean_absolute_error: 0.4418
Epoch 190/200
856/856 [==============================] - 0s 150us/sample - loss: 0.4720 - mean_squared_error: 0.4720 - mean_absolute_error: 0.5299 - val_loss: 0.3241 - val_mean_squared_error: 0.3241 - val_mean_absolute_error: 0.4448
Epoch 191/200
856/856 [==============================] - 0s 148us/sample - loss: 0.4611 - mean_squared_error: 0.4611 - mean_absolute_error: 0.5289 - val_loss: 0.3253 - val_mean_squared_error: 0.3253 - val_mean_absolute_error: 0.4455
Epoch 192/200
856/856 [==============================] - 0s 147us/sample - loss: 0.4795 - mean_squared_error: 0.4795 - mean_absolute_error: 0.5296 - val_loss: 0.3297 - val_mean_squared_error: 0.3297 - val_mean_absolute_error: 0.4478
Epoch 193/200
856/856 [==============================] - 0s 137us/sample - loss: 0.4742 - mean_squared_error: 0.4742 - mean_absolute_error: 0.5342 - val_loss: 0.3204 - val_mean_squared_error: 0.3204 - val_mean_absolute_error: 0.4410
Epoch 194/200
856/856 [==============================] - 0s 146us/sample - loss: 0.4808 - mean_squared_error: 0.4808 - mean_absolute_error: 0.5291 - val_loss: 0.3218 - val_mean_squared_error: 0.3218 - val_mean_absolute_error: 0.4442
Epoch 195/200
856/856 [==============================] - 0s 153us/sample - loss: 0.4609 - mean_squared_error: 0.4609 - mean_absolute_error: 0.5247 - val_loss: 0.3183 - val_mean_squared_error: 0.3183 - val_mean_absolute_error: 0.4367
Epoch 196/200
856/856 [==============================] - 0s 150us/sample - loss: 0.4820 - mean_squared_error: 0.4820 - mean_absolute_error: 0.5351 - val_loss: 0.3202 - val_mean_squared_error: 0.3202 - val_mean_absolute_error: 0.4379
Epoch 197/200
856/856 [==============================] - 0s 146us/sample - loss: 0.4690 - mean_squared_error: 0.4690 - mean_absolute_error: 0.5253 - val_loss: 0.3217 - val_mean_squared_error: 0.3217 - val_mean_absolute_error: 0.4405
Epoch 198/200
856/856 [==============================] - 0s 152us/sample - loss: 0.4554 - mean_squared_error: 0.4554 - mean_absolute_error: 0.5197 - val_loss: 0.3288 - val_mean_squared_error: 0.3288 - val_mean_absolute_error: 0.4451
Epoch 199/200
856/856 [==============================] - 0s 146us/sample - loss: 0.4575 - mean_squared_error: 0.4575 - mean_absolute_error: 0.5196 - val_loss: 0.3200 - val_mean_squared_error: 0.3200 - val_mean_absolute_error: 0.4375
Epoch 200/200
856/856 [==============================] - 0s 148us/sample - loss: 0.4698 - mean_squared_error: 0.4698 - mean_absolute_error: 0.5225 - val_loss: 0.3203 - val_mean_squared_error: 0.3203 - val_mean_absolute_error: 0.4400
    In [317]:
    print(history.history.keys())
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('model loss')
plt.ylabel('loss')
plt.xlabel('epoch')
plt.legend(['train', 'validation'], loc='upper left')
plt.show()

    dict_keys(['loss', 'mean_squared_error', 'mean_absolute_error', 'val_loss', 'val_mean_squared_error', 'val_mean_absolute_error'])
    In [318]:
    y_pred = model1.predict(X_test)
mean_squared_error(y_test, y_pred)
    Out[318]:
    0.4065575532023555

    Observations:The model above has shown some improvement in the accuracy when compared with the pca model above

    In [319]:
    model.save("ModelTrainedOnPC.h5")
    In [320]:
    model1.save("Model.h5")
    In [ ]: